When it first imploded I was on HN taking the position that the astronomical losses were account of marking to market illiquid unicorn poop, and that the real losses to customers would be substantially lower. Eventually some commenters came out of the woodwork to claim some real losses. So I can't say that I've heard of literally no private parties that had losses anymore, it still continues to seem dubious to me-- the entire set of claims around SBF and FTX were obviously implausible even starting from the ridiculous arbitrage gain claims.

If you're looking for a specific thread to pull on, I would suggest the mobilecoin transaction: FTX supposedly lost about a billion dollars to an undisclosed party that deposited an astronomical amount of mobilecoin then received a huge loan on the basis of the deposit (marked to market on a thin market which had suddenly surged) and withdrew the loaned funds, leaving FTX holding a bunch of relatively worthless mobilecoin. Shortly after this was made public one of the mobilecoin executives was murdered in San Francisco, though current reports suggest that this premeditated murder was somehow the result of a domestic dispute.

You would think that recovering these funds would be a top priority as its a substantial and fairly atomic loss which was the product of a pretty obviously fraudulent trade, but it's been total silence on that front from what I've seen.

Even by the low standards of scamcoin casinos, letting someone deposit almost the entire circulating supply of some scamcoin then giving them a withdrawable loan based on the last trade price would be uniquely and surprisingly stupid, and utterly smacks of a laundering scheme.

Of course, one would need to come up with a theory as to why someone (say, an intelligence agency as you suggest) would want to launder a billion dollars to parties connected to mobilecoin... It would have to be an operation more significant than paying RSA to use DualEC in bsafe, something as significant as CryptoAG to justify that kind of payment.