The team supposedly reviews proposals when they reach about 45% commitment to make sure they really are appropriate and aren't duplicates. If that's true, they probably should have looked at this one in more depth. The statistics and the opinion of the majority of the experts support a single site for Astronomy rather than two.
Assuming cosmology is on-topic here (based on this), there are 3 large tags (>200 questions) on Physics SE which would be included, namely [astronomy], [astrophysics], and [cosmology]. Between those three, there are 1388 questions. I am ignoring [general-relativity] and [black-holes] because in my experience the experts on these subjects are more commonly described as physicists than astronomers, but if those are included it bumps the level up to 2579 questions. There are other smaller tags as well such as [observational-astronomy], but let's only consider these big three for now, since they are enough to prove my point. Between them, thir 1388 questions is roughly 7% of all questions on the site. Equivalently, that's roughly 3 questions per day, which is already more than many (roughly 30) of the smaller sites on the network. There is already a thriving community of astronomy experts on Physics SE.
For comparison, if we scaled Physics SE up to Stack Overflow, that would correspond to [android] breaking off and making their own site for Android developers. Equivalently, it would also correspond to roughly 1.4 times the size of [python], a site which was proposed and subsequently closed on Area51 as a duplicate of SO at least once. I'm pretty sure a Python site or an Android developers site is never going to happen, because it's a terrible idea, but that's what would happen here to Physics SE if this site were created. Or we could compare to GitHub. While the tag for GitHub on SO has only about 0.1% of all the questions, the GitHub proposal was closed as a duplicate of SO. Astronomy and Astrophysics are 70 times more important to Physics.SE than GitHub is to SO. Also, keep in mind that the Physics SE community is much smaller than that of SO. On Physics SE, the community and the site is small enough to the point that losing 7% of the questions would be a very significant problem in terms of the prolonged longevity of the site, while on SO it would be barely noticeable. Splitting Astronomy between two sites would have a significant adverse effect on the existing Physics SE community.
Let's look at the existing [astronomy], [astrophysics], and [cosmology] questions on Physics SE in more detail. Of the 1388 questions, 94 are unanswered. That's nearly a 93% answer rate, 5% better than the overall answer rate at Physics SE. It's probably reasonable to expect that cosmology will be less popular than the other two on a site called Astronomy & Astrophysics. If we only consider [astronomy] and [astrophysics], there are 839 questions, of which only 43 are unanswered. That's over a 95% answer rate. By all standards, that's an excellent answer rate. So it's not like the astronomy questions are going unanswered on Physics SE.
Skeptics might claim that this doesn't imply that they're being answered well, but the evidence seems to suggest that they are high quality and highly valued contributions. This query (run on June 12th, but the results haven't changed a lot) shows that among tags with at least 1000 answers, [astronomy] has the highest average answer score. [astronomy]'s 4.02 average votes per answer is significantly higher than [mathematical-physics] at 3.28, [soft-question] at 3.22, and [particle-physics] at 3.15. No other tags are above 3. Going down to tags with 500 questions with this query, the only tag with higher score than [astronomy] is [research-level] which is mostly kept around for historical reasons since it was added to all questions from the closed Theoretical Physics beta site. [astrophysics] comes in at #5 on this tally with 3.29 average votes (above [mathematical-physics]), while [cosmology] is right around the average site-wide. That isn't what I'd expect if astronomy and astrophysics were lowly valued on the site; rather it's indicative that astronomy/astrophysics answers are considered high quality and interesting by many of the existing Physics SE users.
The evidence is that astronomy and astrophysics questions are getting asked in large amounts and answered well on Physics.SE. The people supporting this proposal might claim that the questions on Physics.SE are not representative of the types of questions which they want to ask, and that they would be focusing on more specialized astronomy questions which would not get good answers without being on a specialized site. So let's look at their definition questions. The situation is exactly the opposite of this. The majority are pop-science level questions. They aren't questions that would require an expert in astronomy to answer (for my purposes, an expert is someone at least at the graduate level in terms of knowledge, e.g. the "Expert" and "Academic" categories). Many of these questions could be answered by a simple Google search or by reading the relevant Wikipedia article. There are a few reasonably advanced questions, but nothing so advanced that they probably wouldn't get answered on Physics SE (and, in fact, many of the questions have already been asked on Physics SE). I'm certainly not an expert in astronomy, but I found I immediately knew the answer to 31 of the 40 example questions with only basic high-school level material. I think everyone will agree that, looking at Physics SE's list of recent questions right now, the majority of the questions are at least at the level of an introductory college course in physics. So in a sense it is true that their questions would be atypical for Physics SE, in that they'd be disproportionately bottom-of-the-barrel stuff. However, they're not anywhere near off-topic, and they'd still probably get answers so long as they're asked well and at least slightly less trivial to answer than linking to Wikipedia. This site will not help anyone get an answer to their astronomy questions who couldn't already get it on Physics SE.
The fact that this site is likely to be populated by pop science rather than real science is further evidenced by looking at exactly who has committed. Of the 151 current commitments, only 37 (22%) were self-described as either "Expert" or "Academic or Research-level student". The remaining 78% of the commitments come from people who are at best enthusiastic about the subject, and in many cases just curious. That isn't a very good mix. For contrast, the Biology proposal had 61% in the two Expert categories and Chemistry had 51% (Physics entered beta before committer roles existed). Judging by avid user counts, Chemistry is still arguably having problems with their lack of experts, and this site would have less than half as many self-proclaimed experts. This might not damning evidence by itself, but at the very least it's worrying that the experts (including those already on Physics SE) aren't committing to this proposal. It's very likely that it will only fragment the existing astronomy and astrophysics communities on Physics SE rather than creating a new expert community even if it does succeed. At the moment it doesn't even seem to be creating an expert community at all.
The fact of the matter is that it's at least as common for Astronomy and Physics to be in the same department as separate in Universities. That is to say, at the expert level (which is the level at which SE communities are built), astronomy isn't really separate from the existing physics community any more than geophysics or plasma physics is (two areas which also are sometimes separate and sometimes integrated into physics departments). I assume if I were to propose a site for Plasma Physics, it would be quickly identified as a duplicate, and this case is really no different. In academic culture, astronomy is already fairly well integrated into physics, and since the experts are academics, that's the culture we should care about. The fact that amateurs see these as more disjoint than they really are should not factor into the equation at all.
With all of this, I can not find anyone on Physics.SE who thinks this proposal is a good idea and has a strong reason why. Most people think it is a terrible idea, as evidenced by this thread. That includes professional astronomers and astrophysicists, the very experts who are targeted by this proposal. I was only able to find 7 committers to this proposal who have at least 500 rep on Physics SE. That strongly suggests that the majority of the people here just don't want to participate in Physics SE for whatever reason. I do not know why they are unsatisfied with Physics SE, but unless they voice a reason there I also do not think it is important why. We've already established that there's a sizable number of people on Physics SE who are doing astrophysics; the inability of this proposal to attract many of them speaks volumes.
While I think I've written too much already, it's worth going over the checklist from the "Should my idea be part of an existing site, or its own site?" section of the FAQ one more time to see what SE's official decision should have been:
In general, if a site makes sense as part of a bigger site, it's better to have one big site than a bunch of little niche sites. Site X should be subsumed by site Y if:
- Almost all X questions are on-topic for site Y
- If Y already exists, it already has a tag for X, and nobody is complaining
- You're not creating such a big group that you don't have enough experts to answer all possible questions
- There's a high probability that users of site Y would enjoy seeing the occasional question about X
- Yes, all questions here would be on topic on Physics SE. There is no debate on this point.
- Yes, several tags already exist, are getting good answers, and none of the experts on Physics SE are complaining about this. The people who are complaining don't even participate on the Physics site and don't have a good reason why.
- Yes, this proposal has a dearth of experts. The experts are already on Physics SE and don't want to split. The people who want a new site are almost exclusively novices who could get their answers on Physics SE, Wikipedia, Google, Yahoo! Answers, or pretty much anywhere else.
- Yes, there is interest in Astronomy on Physics SE. The answers in [astronomy] and [astrophysics] tend to generate significantly more votes than average, signifying strong community interest.
On all four counts, Physics SE should subsume this proposal. This is exactly the sort of case which closing as duplicates is supposed to cover (according to the blog posts Merging Season and An Area 51 Apology — and Clarification). My only guess as for why it wasn't closed is that the team lacks anyone knowledgeable about the culture of physics and the Physics site, and thought this might be worth a try. It's not worth a try, any more than a Python-specific or Java-specific site is worth a try. There are only two possible outcomes. Either the site will fail, in which case it will get folded back into Physics SE (but not without hurt feelings and a lot of work on the part of everyone involved), or it will succeed and divide the community, creating a new site filled mostly with pop-science and few experts, benefiting precisely nobody.