This is the first time I've seen a high profile youtube channel actually discuss the 'metagame' of competition and repertoire selection.
You can definitely see the competitors are aware of the trends. Scriabin sonatas have gotten very popular. There's an awareness that pieces like Ravel Valse, Apassionata, Petrushka are overplayed and get avoided, and then one person who actually picks it gets to be the only one to play them. Competitors are also aware of what pieces always lead to getting eliminated like that Franck Chorale piece, so they become 'statement' or gauntlet pieces. You're telling the jury 'I know this piece has a terrible hit rate in competitions. I'm picking it anyway because I'm that good.' High risk but possibly high reward.
Since we know Gaspard/Petrushka/Bach-Busoni Chaconne/Hammerklavier will be picked by multiple people, and if you're the megavirtuoso, the legitimate main character of the story, you can intentionally pick them knowing you're going to outclass everyone. I'm pretty sure Aristo Sham was at least aware of this for his early solo rounds.
And I've seen this trend of disfavored showpieces. Daniil Trifonov, Dimitri Shiskin, Alexander Lubyantsev permanently killed Mephisto Waltz with insane performances, at various Tchaikovsky competitions. Breakneck 10 minute 0 sec renditions, no safety, no cowardice whatsoever. You know everyone will compare you to that standard, so if you can't play it like them you're better off not programming it.
Islamey you'd be playing solely as a virtuosic display but it may not be difficult enough at this caliber of competition. If I ask, 'where are you worried that a disaster will happen' for Schumann fantasie, Scriabin no 5, Petrushka, or Don Juan', everyone will tell you oh it's the coda, it's the ridiculous leaps, etc. But there's no such spot for Islamey. Ergo it's not difficult enough given its scale and not 'musically serious' enough to make up for it?
It's interesting why the ultra-difficult region of repertoire is so limited. I think pieces like Schumann Toccata or Liszt Lucrezie de Borgia or Liszt-Paganini s140 original version or Liszt Spanish Fantasy are so risky and require such warmup, and no one wants to make the first move without seeing if the jury will reward them. (I am alluding to Yi Chung Huang, but from smaller competition outings, it seems like his juries don't like ultravirtuosic music outside the established core of Gaspard/Don Juan/Petrushka/Brahms-Paganini). Maybe too gratuitously difficult and not musically substantive enough.
And the nature of these difficulties are so pernicious and require so much short-term memory reinforcement, you still 4 hours of other music to keep ready at the same time. My theory is if they ever run smaller competitions with maybe 2 hours of music required tops, you'd see these exotic virtuoso works programmed more. But, every big competition is obsessed with requiring 4, 5 hours of music total. I kind of want to see what these pianists can do with an ultra-polished, super high conviction and passionate 2 hours of their best music, and not spreading themselves too thin.