In the Rittenhouse case, both prosecution and defence agree that Rittenhouse was a person who had a gun in his arms and fired it, resulting in death. The thing the prosecution needed to prove is that in doing so, he committed a crime. Prior to the trial, it was not necessary to express agnosticism about whether he had a firearm, fired it, or firing it resulted in anyone's death, even if there was some sense in which he was "innocent until proven guilty" of the conduct being criminal.
In this case, there is essentially no room for dispute that stealing $71 million, engaging in a vast money laundering conspiracy is in fact illegal. If these people actually had these accounts and actually used this money in this manner, there is no chance they will not be found guilty. The affidavit is not compatible with a set of the same basic facts leading to a different legal conclusion.
So the question is whether or not you think there's any possibility that the feds cavalierly misidentified the people in possession of these accounts. That seems pretty unlikely, given the affidavit suggests withdrawing small amounts of the stolen money to use Uber under their actual names, buy stuff on PlayStation under their actual names, etc; and that the private keys were taken from a cloud storage account actually belonging to the guy; and that the woman contacted various exchanges and talked about the actual company she actually owns, and the guy actually has a documented record of talking extensively about cryptocurrency including on this site using his actual name.
It is, of course, possible that the entire affidavit is a lie, made up whole cloth and all of this evidence is totally fake and the accused were minding their own business working on their gourmet cupcake business in Kansas City, and they don't understand anything about no crypto-whatsit. But I don't think that's a scenario that really requires much investigation, and it instead is a level of solipsism on par with "we can't actually know if gravity will cause us to fall" or "what if this is a simulation".
What I'm saying is that not all uncertainty is equal, either in kind or in probability, and so it makes more sense to be honest about that than equivocate across very dissimilar cases.