Cool post.
Refuting 10...a5 (Najdorf - English attack)

I think the problem for black is after 10...a5 11.a4 the queenside becomes more closed than if black played 10...Nbd7 followed by 11...b5 and 12...b4.
Since in the English attack black typically wants to play on the queenside, closing the queenside should favor white.

I believe Black is fine and the move is still played in GM level(obviously OP understands more than Maxim Vachier Lagrave and Topalov) but if logozar(or should I say logozar's engine?) says it's refuted , who am I to disagree.
I'm not considering switching for no reason...
Anyways if you want me to explain how I figured this out:
I played a lot of practice games with German_MagnusCarlsen. He was pretty booked up as white in the English attack, so I started learning more theory on it myself. This lead me down the line of Nh5 and g6 which GM Rafael had recommended in his video series and on the repertoire on chessable (Which I published, but I hadn't studied the whole thing).
I had some success playing it as black. Eventually, I played a NM who after I won multiple games against he booked up. Against what he played, I could not find equality following GM Rafael's recommendations.
I analyzed with a computer, it agreed. Black isn't equal in that line.
It showed an interesting queen sac (positional - for white) that looked interesting, and it seemed to be dynamically equal.
Unfortunately for black, white can improve earlier.
I decided to reveal the analysis in the forum but doubted that people would take me seriously if I only analyzed a little bit (one line within it) so I set out to find the refutation to the major alternatives for black throughout the line (after 10...a5).
Yes, I used a computer.
If there is any way for black to equalize after 10...a5 in this line, I'd like to know!

How white can get an advantage after 15...f5(!) 16.g6 hxg6 17.Rxg6 Rf6? I think Black is OK after the rook retreat with ...f4 and ...Rh6.
The critical line is 16.gxf6 e.p. Rxf6 17.Bg5 Rxf3 18.Bxe7 Qxe7 19.Qxd6, and the ending is unpleasant for Black.
Thanks, I edited my post to include your analysis.

I assume the improvement is 15...f5 16. g6 hxg6 as pfren has already mentioned
He also mentioned that 16.gxf6 e.p. leads to an unpleasant endgame for black.

I'm not sure the best way to phrase it. What I meant by refuted is it leads to positions by force that are from += to +/-

Instead of saying "refuted" it would be more accurate to say that you have no idea what to do after 10...a5 11.a4 (and analysing with the engine made things worst as always is the case in such complicated positions).
I had quite a bit of prep, but now I think white is better if he is accurate.
(Prep to up to 17...Nf4 18.Bxf4 exf4 19.Nd4 d5) (19.Qd2 is an improvement)
from GM Rafael Leitao's video series.
It's not that I have no idea what to do, I just think that objectively black is worse in the a4 line if white plays correctly.
Perhaps there is some idea/line that equalizes in it that I'm not familiar with.

And jengaias, thanks :)
The list of GM's who play it might come in handy, I might want to look up some of the games they played in it.

Pfren, is there any way for black to equalize in the 11.a4 line?
Clearly 15...f5 16.exf6 is quite unpleasant for black.
Where can black improve?

I think Na6 will transpose into Bb5 lines if they play Bb5 and otherwise might transpose into the lines I don't think are good for black is he plays Nb4.
Though the idea of playing Na6-Nc7 in the a4 lines seems interesting, I'd like to know if this works.

No cowards allowed. PLAY NBD7 AND WIN LIKE A MAN!
Why dont you play Nbd7? Black is usually not getting mated there.
10...a5 in the English attack (Najdorf)
The second most common move which has been played at very least hundreds of times at master level.