People went to that, myself included because I feel that that majority of the suggested solutions in your OP wouldn't solve the issue given, maybe even make it worse. Play time is not an indicator of any skill or knowledge beyond the first few experiences (which are well defined in the requirements of the mentor system already), and I disagree that it increases probability of being useful beyond a nominal amount after first experiences (20 hours compared to 100 hours in, the understanding of basics changes incredibly little if at all). You don't know what they've done with that playtime - there is an incredible amount of things to do and not do, playtime would dictate what.. how well they can afk could be one yes lol.
I do not believe playtime should be seriously considered as it will only reduce the amount of mentors we have while offering near zero quality improvement (if not actually a less). Besides the 1,000 instance requirement assures that people have seen basic mechanics, adding time that would go over the amount of time it takes to do 1,000 instances, MSQ, level 60s, wouldn't add extra basic knowledge if you didn't catch it already in that time frame (seriously, after all that content and you don't have basics down? Extra time isn't going to solve it lol). Time would be a gate without results.. You do not need to play for very long at all to learn how to dodge circles, and if that's your only gripe (basic mechanics) then again playtime is the antithesis for a solution.
At 1 hour into the game I can guarantee you, on any mmorpg I've got the basics (dodging the bad circles, pressing 1,2,3) - and I assure you these people who generally "get" a game do too. But you'd gate these people because you want to assure better quality yes? But if you check lodestone (as I've done quite a bit) a lot of these people who mess up have other high end classes or achievements, again you'll not do much with playtime. I guess just play some LoL/DotA/HoTs for a while and you'll see easily people who have hundreds of hours doing awful and those with a tenth doing better, play time is not an indicator of skill beyond the first hours (at which point you either picked up the basics or you didn't).
If you want to weed out poor quality the only good way to do that is a test - specifically a test that relies only on yourself, or I suppose we could just leave the settings alone :/. So this is where the hang up arises as I've seen your issues myself, but I don't think you're ever going to be able to solve them unless you introduce an actual quality test. Also as I'm keenly aware of the issue Atoli was bringing up and like the idea of tests as you could complete these very quickly (given you're skill level, and the test level you have to achieve could be based on the rank you want - basic mentor doesn't even need to do them I suggest, just a roulette mentor and higher requirements for higher tier mentors). I don't think you can solve mentors being rude unfortunately, even if you can make sure whatever tier/title they yield means something.
We don't need general helpers (those kind people as Atoli described) having a harder time being helpful to the community, again I am strongly against the suggestion of just upping quantity rules because at 1,000 to 2,000 will mean no substantial skill difference.. just a lot of time and a lot less mentors. General tier helpers do not need any more gates, except to turn off the mentor titles they didn't unlock (like DoH/DoL unlocker being able to mentor as a DoW, but didn't meet the DoW qualification). If you want better helpers in combat though then again, just adding quantity isn't going to amount to anything besides seeing less mentors. Perhaps that's a bonus? Of course as talked with Atoli, doesn't really make sense to require massively difficult content to help people go kill HM titan, although I argue a bit more so on EX content, but it might not hurt to add future tiers to the system who want to be mentors of end game (in which case you do have to do something quite difficult). Also I believe it would help to show those who are general helpers and those who are particularly good at their job, since you can have someone be helpful in a dungeon but not actually know how to play their job that well and shouldn't be watched as a skilled player on that job (would help to know, might reduce mentor hate as well "oh they're new to BLM, but they can make sure we get through the dungeon at least!". As the worst BLM I've seen in recent memory was actually a mentor, made me cry inside lol (but they were at least leveling, so hopefully it just meant "learning")).
1,000 instances at 10 minutes a pop is 166 hours (10 minutes is an estimate, but we do have many other requirements too so I feel its fair). If you don't know the basics at 166 hours of actual play time you're not going to know them at 200 (and a generic play time requirement will include AFK time, which many people just leave their character on all the time - so unless you seriously inflate the playtime number.. it's again for multiple reasons, not going to be helpful). Only those who had to learn and prove that they learned the basics would know them for sure, and you do that by a test.
The rate at which people improve exponentially decreases with an increase in quantity is what I'm saying. The difference from 1 hour to 20 might be a lot, 20 to 100 less so - especially for "basic" mechanics. The further out you get the more meaningless the change is. As I said if you're awful with basic mechanics at 1,000 there is not a significant change you'll actually be good at 2,000. Except that the novice chat room mentors probably just got cut in half or more.
I guess i'm being a little vehement on this time thing lol but I seriously value those who take the time to teach and help others and I loved reading Atoli (and others) talking about helping even though they haven't played a boat load and or mastered everything with perfect rotations on savage. Extra time gates then we have will only hurt the kind. Simultaneously I'd love to see a solution to your issue where we can have certain mentors expected to be decently good at the content you see them flagged as mentor in (in which case, time isn't going to change it, nor will achievements). Also why I've been trying to suggest general mentor, and "star"/ or basically proficiencies inside the mentor system as well as it turning off when you're in content you don't meet the "advanced" requirements to be mentor for (like end game content).
If you have to pass a content that says you know how to do basic stuff (something that specifically only relies on you, no carrying team mates) then you can have 100% expectation that they can and are capable of doing it, similarly they the mentor know that is the expectation of them (as it was their test to gain the rank whatever they earned).