Because a lot of beginners joined recently, it makes sense that your percentile rises.
Rating Deflation

lol, at 97.2% percentile
Yes, this is what I am talking about. Your rapid percentile is 97.2%, but your rating is 1521. Both the elo and glicko system, if they are working properly, more or less require a 97.2% performance to map to around 2000 rating. So something is wrong.

ratings are turbo inflated here
But they aren't. My OTB rating was 1299 last I played. In ye olden days, my "slow" rating on chess.com was over 1400, and you could say it was a bit inflated. Now it's down to 1280. As I mentioned, my percentile is rising (possibly due to new players), but also my accuracy (as judged against the computer) is rising too.
What's more, I have played enough to judge various situations. Against some (weaker) players, I am strategically outplaying them, until generallly they make a tactical mistake, but even if they don't make the tactical mistake they lose strategically. Then, against some (stronger) players, the opposite happens--eventually I make a tactical mistake, but even if I don't make one they win strategically. Finally, against some (equal) players, the battle goes on "in the variations," where various things could happen, but they don't, because both sides block the other's threats as they happen--until eventually someone misses something.
All I'm saying is, the level of play I used to see from 1400+ players I'm now seeing from 1200ish players. Yesterday I had an 1195 player take three moves to reposition a knight. As soon as he made the first move I could see what he was doing, and that it was correct to do it, and that I couldn't really stop it. In ye olden days, no 1195 would do that--more 1400+ to do it.

So...my current rapid rating continues to fall and is now 1280.
My "percentile" continues to rise and is now 92.9%, so supposedly I'm rated higher than that amount of the population. It's actually a bit hard to believe.
But either way, isn't it a huge mismatch?
It just means more and more new players who can't play at all.
I don't doubt that, except the rating falling at the same time. I don't care from an ego perspective. It just makes it hard to track progress, or lack of progress.

why?
the system would work perfectly well if everyone got 4000 rating boost, all the presets got a 4000 boost, and the minimum rating allowed was boosted by 4000

why?
the system would work perfectly well if everyone got 4000 rating boost, all the presets got a 4000 boost, and the minimum rating allowed was boosted by 4000
That's an idea, but I don't think it works.
A better improvement would be to have the distribution be symmterical around a fixed mean. Traditionally--for elo and glicko--that mean would be 1200. But if we are trying to line up with USCF--FIDE is rather pointless to try--then you might decide the mean should be 800 instead. The current mean of barely over 600, and especially the very strong rightward skew is an indication that something is clearly broken. And even more than those things, I am pretty sure the mean is moving.

Or to put it another way, this is not a healthy glicko distribution.
Agreed. An average of 600 is horrible. I created this thread about it: https://www.chess.com/forum/view/general/why-is-the-50th-percentile-on-this-site-600
Same question I asked arjunjagen...if you don't mind...do you have a USCF rating and would you mind saying what it is?
So...my current rapid rating continues to fall and is now 1280.
My "percentile" continues to rise and is now 92.9%, so supposedly I'm rated higher than that amount of the population. It's actually a bit hard to believe.
But either way, isn't it a huge mismatch?