By paragraph:
1) Glad you said field matters. Several where LOR can trump even 15 points on boards gap as long as they're past the cutoff threshold. And PLENTY where research=/=volunteer work. Rad onc for one that's basically indisputable? Or however you spell it?
2) I don't think they care about this nearly as much as the fact that if you haven't worked with research and they are a residency requiring research (which may be speciality dependent as you do say), you'll have to take 10 hours just to figure out how to get Adobe PS to do all you need. You'll need 10 hours just to figure out how to get Illustrator to crop and line things up under correct settings etc. You'll need 10 hrs for prisim. You'll need 10 hrs for freeware R software or SPSS. You'll constantly be pissed because you worked but can't publish all the tons of negative data that took longer than the few bits of great positive data, even if its useful negative data (you won't be used to this idea if you've never done research, particularly basic). You'll need to mess up a few PCR runs and blots. Staff will need to take an intensive 10 hours (not just you but also a facility manager) teaching you how to use confocal microscopy equipment and software, cryostat/vibrotome (if you're lab uses them, if not, something else like radiation administration etc etc etc). You'll need intensive help getting your writing and organization of an initial manuscript draft together, not only so its quality data and information but so its in the scientific vernacular that I swear is sometimes still very tough for me (I've done some serious full time research after bachelors). They'll have to walk you through submission and re-submission usually; probably if its basic research they'll have to let you know when you have enough to actually submit (fun little thing someone sent me:
http://www.phdcomics.com/comics/archive.php?comicid=1382). More than anything if someone isn't used to organizing the HUGE amount of info even a lazy member of lab accrues over months and months, you'll do a lot of work and probably not have a systematic way of recording everything thus lots of things have to be re-done or samples thrown out (even clinical but more a basic science problem). Bottom line: they don't want a scumbag but more so than character/self-growth they do NOT want to have to invest time training you to have a research bearing to meet program research requirements. It is a pain in the *** to have people that will require an extra 500+ hours of man-hours from full time staff to get you up to par with keeping good data, using all the usual equipment, etc.
3) Some programs don't care. They put out work others would consider 'trash' all the time and meet requirements and chug right along. If you did nothing but take old patient samples and run elizas and report quantitative data for 8 publications that you basically did the same thing on, just for different drugs/proteins of interest, I'd say that was complete 'trash.' But in some circles that would seem more merit worthy (again depending on what they're looking to have you do) than all I've done for a few years and LOTS of late nights, etc etc...
4) I've heard this and somewhat agree. It ties in with above. Hospital wants to tell lay public they support research as soon as they click into the website, and no one usually calls out an entire program even though there are some notorious 'trash' research institutes (sortof like grade inflation I guess?), so they'd take the guy with the eliza background, keep 'em doing that and taking the crappier call after a year or so rather than a candidate who is genuinely interested in some very hard regen. medicine unless the person is such a wrecking ball that they bring lots of academic cash into the residency (super rare I'd have to think). Other programs, however, and there is some very hard research going into vascular/GS inquiry some places, have different emphases and from where I'm from I can tell you for certain senior people say that basic research is much better 'respected' because you have to muddle through/troubleshoot a lot more for a complete product. On the other hand one big guy who does hold this opinion also thinks it's BS (says, [paraphrased] "why should we have fellows doing genetics of drosophila? They are only doing it to go private and we all know it. But if they don't have the basic science in [anonymous speciality], then they don't get a job; at one point people stuck with clinical but then someone on the block did some basic research and that was distinctive, so then everyone on the block is holding up overly academic basic research positions. Problem being that not only would they be more useful for production if they focused on other things, but that the people who really love research and academics have to compete with people who only want to do private but will pretend along and are also very intelligent and highly motivated; 'we' [they] train people at not huge expense but great effort and time who then leave the field now.")
I will say that I completely agree that merit demands quality>quantity but that sometimes programs aren't even emphasizing quality so quantity might be the way to go for 'success' within the placement system. I'd also point out that charting outcomes includes truly weak research submissions (abstracts or poster presentations at your annual school student research poster day for projects about on par with the 100 elizas and publish model I described) as a 'publication' so don't flip out if you see your intended speciality has an average match statistic of 10 papers, etc.
Wow this was way too long, even to proof/edit. Condolences to whoever makes it to the end. Truly sorry I got out of hand.