I think distributed computing of large tablebases (large=8-9 men, at the time of writing) is technically possible. But this going on in real life I only see it happening if egtb generation were somehow how a by-product of time-space-proving crypto-currency mining. I am thinking in particular of Chia crypto and the current run on large capacity SSDs and hard drives. Discover a way to have tablebase generation become a way to make money in China, and the computers will surely follow.
You've been pondering endgame generation for a very long time -- author of the 2001 Koistinen algorithm sketch, after all. I bet you've noticed that when a newbie comes on board the top questions are:
1. Where can I find/download/install tablebases?
2. Why don't you just generate and save the important positions?
3. Why don't you use distributed computing like folding@home?
A few experts then respond, sometimes patiently, sometimes not.
Oh, let us not forget four,
4. Yeah, that's nice, but when are you going to do the N+1 tables.
c.f.
viewtopic.php?f=6&t=7618 for the eternal answer "never". With the implied eternal words of wisdom: never say never. (Pertaining to extrapolation of Kryder's Law, we do not yet have somewhat affordable 10 TB SSDs.)
I divide the main idea into five aspects.
a) distributed computing
b) distributed storage
c) distributed probing
d) project management
e) participant motivation
For a) I think the contributors necessarily divide into two camps. There will be a small group of participants with high-capacity computers, defined as able to generate pawnless 8-man tables in isolation. Then there is a larger group of ordinary computers owning regular desktops and laptop. The work allocation manager calls upon them to compute and make available pawn slices.
Items b) and c) scare me personally. I do not know high-security programming and would be hesitant to even adopt an existing infrastructure. The thought of the system being hijacked to work as a spam bot network, or DDOS net, or storing large videos of a "certain nature", or worst of all injecting malware onto peoples computers -- that's a though I don't want weighing on my conscience. A more expensive alternative is for participants buy-in to one of the major cloud providers such Google, Microsoft, or Amazon; all the hosting is on rented computers.
d) The project won't just run itself. It takes plenty of energy to be one of the organizers.
e) And what I believe to be the biggest show-stopper: what would motivate people to participate in droves? The stockfish and lczero training networks are impressive, but that runs only among the dozens to hundreds of contributors. What takes it to the tens of thousands and beyond? Endgame crypto-mining is my answer, if it were even possible. Outside that, how would it become a fashion?