OP thinks that people will actually get above 1400 and then stay there...lol.
OP also thinks that decay is the only factor that will cause people to drop below 1250.

There's plenty of bad players in the game. They'll drop below 1200.

If people are also losing and winning at 1200 and averaging out to around 1200 in a mass population, then that can still make it the top 25%.

Not to mention all of the dodging.

While Elo decay only effects people above 1400, it still has the effect of pushing the AVERAGE downwards as points leak out of the system.

I will disagree with the way you've framed the quoted statement -- specifically the idea that '1200 = top 25%'. It's true that the average is below 1200 amongst ranked players, I just don't think it's WELL below 1200.

That being said, the 25% stat is likely correct from a different angle -- the angle of "all active league of legends players"

Not all level 30s play ranked, and not all players are level 30. So, if you are playing ranked at all, you are already part of a smaller, more competitive group of players. I think it's fair to say that a 1200 ranked elo player IS top 25% of LoL players for this reason, maybe even more elite than that. As an example, if you assumed only half of players are 30, and of those, only half play ranked, any score over 1200 would be something like the top 12% of LoL players. I don't know the actual numbers off hand, but a lot of people aren't level 30, and a lot of people do not play ranked.

This is the same for any game. Some people play to compete, some do not. Look at WoW -- if you have a 'bad' arena team, you are already a really good PVPer -- in the context of the entire population of people who play WoW.

So in summary -- Is ranked elo 1200 top 25% of all active LoL players? Certainly. Top 25% of ranked players? No.

welp... i'm in the top 40,000 @ 1472 ELO... Assuming 25% is 1200, and there are ~250,000 people above 1200..., there are 1 million people playing ranked. (.25*1,000,000=250,000). Does this make any sense? Meh. Maybe. 1 million out of 33 million accounts playing ranked. If thats true, that would put me near the top 4% (40,000/1,000,000=.04*100=4%). What if 1200 is the top lets say 15%?

250,000 (ranked players @ 1200)/.15%=1,666,666 players.in ranked. That would put me at the top 2.4%. Sounds sort of un-realistic. This propogates a RIDICULOUS curve for the % of players at that ELO at ELO gains of even 1 or 2 points. I think 1200 ELO is closer to 50% of all players, because I have a hard time believing i'm a top 5% player or better. This is all guesswork of course and none if this has any validity whatsoever. Much theorycrafting has been done but this is how I rationalize that I am not quite a 5% minority in this game.

riot horey sheet

Well, consider that .03% of high school basketball players eventually go pro. If you are the equivalent of a varsity high school basketball player in LoL, those numbers line up pretty well. I did not scrutinize the specifics of your # of players or whatever-- I actually don't have the stats. But your conclusion seems reasonable generally.