It seems they have rejected initial versions of the paper, since there had been later updates and clarifications based on the reviews. So it seems this was bene

It seems they have rejected initial versions of the paper, since there had been later updates and clarifications based on the reviews. So it seems this was beneficial in the end and how review process should work? Especially since this was groundbreaking work it makes sense there is more effort put into explaining why it works instead of relying too much on good benchmark results.