Joshi Kirti's amended proof based on Mochizuki's earlier 'proof' through Interuniversal Teichmuller theory is correct and will be accepted by Western domain experts as such within 5 years.
How much modification would you allow to still resolve YES? I.e. does it need to be the current (March 15) version of the proof or are later versions allowed? If later versions end up having other co-authors with substantial contributions will it still count for YES?
Also: related, weaker condition, shorter timeframe:
@AIBear it's common for big proofs to contain minor errors or even substantial ones (like Wiles-Taylor).
This is a bit of a judgement call but overall I'd lean towards a fairly loose judgement. If the approach is fundamentally flawed all the editing & revising in the world won't change that.
@AIBear if other coauthors with substantial contributions end up appearing that's probably a No.
(Just like it seems Mochizukis original proof is substantially incomplete at least)