Oh my god I love your questions! I don't know if I can do justice to all of them in a comment thread! I'll do my best, but you've given me essays worth of material here :)
1) You're right that many many authors and fics don't get read a lot, and sampling from rec lists will probably be biased towards already popular fics. But! Kudos on AO3 are not always a good meter of popularity, especially for older fics. (The example I always trot out here is annafugazzi's Bond, which dates from 2006 but which the author only uploaded to AO3 in 2014. However, when she uploaded it there was a months-long gap involving technical difficulties, and she didn't add many tags or a rating, which hurts its visibility. Right now this fic has only 216 kudos on AO3, which as you and I know, is not at ALL indicative of its wider popularity in the fandom or of its true quality).
Still, 216 kudos ain't nothing :) Not like the <5 you mentioned.
So I went and looked at my data again to see if I could give you some more info. Of the 1490 fics I looked at, 911 are available on AO3 (that I know of). As it turns out, I made a note of how many kudos each of those fics had on AO3 at the time I read it or added it to my database. For really popular fics, this number will change dramatically over time. But for less popular fics that have been on AO3 for a while, their kudo numbers stay pretty stable. If we ignore fics published in 2015 (because their kudo numbers will still be highly changeable) I now have 880 fics on AO3 that I can say something about.
If you go as high as 20 kudos, I've got lots more fics on my list (and even a few more with <5 that weren't recced by capitu but that I added myself because they struck my fancy)
So while I absolutely agree with you that my list is not going to contain that many unpopular fics, it still may be true that I have a representative sample anyway. And it's also not always clear how to (objectively) determine how popular a fic is anyway.
I know that's not a complete answer, and you asked many more things, but I'll post this part and answer more of your questions subsequently :)
no subject
1) You're right that many many authors and fics don't get read a lot, and sampling from rec lists will probably be biased towards already popular fics. But! Kudos on AO3 are not always a good meter of popularity, especially for older fics. (The example I always trot out here is
Still, 216 kudos ain't nothing :) Not like the <5 you mentioned.
So I went and looked at my data again to see if I could give you some more info. Of the 1490 fics I looked at, 911 are available on AO3 (that I know of). As it turns out, I made a note of how many kudos each of those fics had on AO3 at the time I read it or added it to my database. For really popular fics, this number will change dramatically over time. But for less popular fics that have been on AO3 for a while, their kudo numbers stay pretty stable. If we ignore fics published in 2015 (because their kudo numbers will still be highly changeable) I now have 880 fics on AO3 that I can say something about.
Of those, I have 3 fics recced by capitu on my list which have less than 5 kudos on AO3 as of today:
archiveofourown.org/works/1950570 (4 kudos, bottom!Draco)
archiveofourown.org/works/1157336 (5 kudos, bottom!Harry)
archiveofourown.org/works/0301315 (5 kudos, switching)
If you go as high as 20 kudos, I've got lots more fics on my list (and even a few more with <5 that weren't recced by capitu but that I added myself because they struck my fancy)
So while I absolutely agree with you that my list is not going to contain that many unpopular fics, it still may be true that I have a representative sample anyway. And it's also not always clear how to (objectively) determine how popular a fic is anyway.
I know that's not a complete answer, and you asked many more things, but I'll post this part and answer more of your questions subsequently :)