(c) I am sympathetic, and it's true that some departmental cultures might not be productive for you. That's a decision and a call you have to make.
(b) I think confidence has little to do with outcomes. A confident, talented scholar will have a better outcome from Michigan than from Wayne State. Sure, departmental differences, living situation, etc. makes a difference, but those differences (I believe) are ones that help you decide between UNC and Wisconsin, or between Yale and Princeton. It's asking too much of "intangibles" to ask them to overcome large differences in program quality.
(a) As a theorist, I am particularly sensitive to this point. My solution is to, broadly, just place those schools which have strong theory departments alongside other top schools--that is, if someone were to ask me (as a theorist) what the top 15 schools were, I'd list Northwestern and John Hopkins and Notre Dame, alongside the top 10-12 overall programs. Notre Dame's excellence in the theory subfield makes it a top program for theorists. Obviously, there's a balance between overall program strength and strength in subfield.
I do want to make several additional points, however. I pointed out earlier that we will all be hired (or not hired) on the basis of perception. Search committees will use heuristics to make their job easier in a high noise low signal environment. Among other heuristics, search committees will think about the overall reputation of your program, of whether they've heard of your work, of the reputation of your advisor and references, etc.--just as is the case in graduate school admissions. It's not fair. In applying to graduate school (and in applying for jobs), you may very well have candidates of identical quality from vastly different backgrounds. But the sad (yet understandable) truth is that in applying to grad school you'll have a better shot with LORs from well-known scholars and a degree from a top institution than a degree from a no-name place and unrecognizable LOR writers. The same is true in getting a job.
Now, where does this come home? As a theorist currently in graduate school, I have to encourage my fellow theorists--nay, I beseech you!--think about the job market. I began regularly cruising the Chronicle's job postings and those on Higher Ed Jobs before I began graduate school. You'll notice, there are almost no positions out there that are hiring just political theorists. That is to say, you will be unlikely to get a job simply based on being a damn good political theorist. Of the many universities and colleges in the U.S., a relatively small proportion of them can support people who just do political theory. Many/most of the jobs require or state preference for theorists who can also teach Con Law, Methods, American Politics, etc. You will probably be hired as a swiss army knife, not as a full-time political theorist. If you are certain that you are in the top 5-10 political theory Ph.D. students in your year, perhaps you will be fine--but otherwise, you need to seek broad training in a variety of areas--because that's what the jobs require.
One last word on perception and the job market: It is unlikely that your search committee will have more than 1 or 2 political theorists on it. It's quite likely, at many institutions, that if they're hiring a political theorist it's because their only one is retiring. Much more likely, your search committee will have Americanists and comparativists. Most people in our discipline (non-theorists) have little ability to accurately judge the quality of political theory work. Thus, the reliance on heuristics becomes even more pronounced. Also, many in our discipline don't know the names of top political theorists (or they might know just a handful of the biggest names). So to an even greater degree, decisions will not be made entirely on the basis of merit or real quality, but on perceived quality of your training, corroborated by LORs and publications.
I have spoken to a number of well-known, tenured political theorists at top schools, and the near unanimous advice I've received follows these lines:
1. go to a school where you can produce AND PUBLISH quality research in your subfield
2. go to a school where you can learn and work outside of political theory (in other words, probably don't go somewhere with a solid theory program and low quality overall program).
3. If you have any ability or inclination, take methods courses so you can contribute to quantitative research agendas, teach methods, and maybe coauthor (even if you don't do quant work on your own).
4. Cultivate good relationships with well known figures.
5. Practice telling non political theorists why your work is important.
Sorry for the novel, and I hope it helps. It's a killer decision--last year, I was choosing between theory boutique programs and some broader programs. There isn't any one-size-fits-all approach. I'm just trying to offer some things to think about.
You're definitely right that there are certain things under an applicant's control - GRE/GPA, SOP, etc. By all means, make your application packet as strong as possible, and use all the resources you have at your disposal to do so.
I wasn't trying to say that the entire process is all chaos and that your chances are completely random. My point was more about how some applicants approach specific programs. An applicant can do certain things to make their application broadly more attractive to all schools (do research work, go to conferences, etc) but when it comes to a particular program, there is no "special sauce" for getting in. A number of users here have been laboring over tactics to get into specific programs, and there's just too much going on to reliably predict whether or not this or that tweak will get you into that program.
In short, most of the applicants on gradcafe are already aware of the general types of things they could do in order to make a more attractive application. All programs, no matter what their rank, are looking for basically the same criteria. It would be an error, however, to attribute one's getting into any one particular program directly to something in their application packet. (e.g. You might have a great application that is worthy of getting into some of the CHYMPS, but there's no guarantee that that packet will get into Harvard in particular).
I've just asked questions about stuff that I'm actually interested in about the program/the classes she's teaching/her research--I don't get the impression my POI is expecting me to confirm my acceptance yet at all either, especially given how early it is and how many schools have not even sent results yet (ahhh yet another reason I love her)
On a related note: This sounds like it should be basic but I’m not sure how to determine the subfield rankings of schools below #10? I saw a thread saying that until #15 or 20 should be “if you have to ask if they are, they aren’t” but I’m not sure if this is great advice to someone like me who tends to second guess everything.