I'm OK with the burden of proof. I was basically saying either assumption (in, or not) is an assumption.....and requires proof.
If there was no indication or evidence whatsoever that mvis IP might be in IVAS (latter version), then maybe you could argue that the better default assumption is NOT in, until proof provided. But given the knowledge of mvis in HL2 (at least early version), and evidence mvis might be inside IVAS (at least early version), it seems like one needs some proof that mvis is NOT inside (latter) IVAS, before claiming the NOT IN assumption is the better assumption. given "onus probandi" applies to either assumption.
I don't see a problem with anyone running with either assumption, if they choose. Either can be defended as a assumption, with associated risk to the chosen assumption.
If one doesn't want to run with an assumption, but rather wait for proof, then it seems prudent to acknowledge the unknown, and don't assume.