I think, Luke argues that by defining how to store arbitrary data, a precedence is made that Bitcoin is for arbitrary data storage, making all other arbitrary data problematic, too.
OP_RETURN was for hashes and then for hashes plus some meta data. With 100kB it's not meta data or hashes anymore.
The problem won't be that you have this and that sequence of bytes on your machine but that you are using Bitcoin period.
So by this logic if CSAM is included in Inscriptions that could now land you in jail because Bitcoin Core increased the OP_RETURN policy default. (Even though it was already possible to do that.)
Do you find this argument convincing?
I think that if Luke and others can see it that way, politicians might parrot the sentiment in defense of laws against Bitcoin. Politicians only need majorities and Luke's arguments did move the needle among Bitcoiners or we wouldn't have this debate.
Then it sounds like you're saying if enough people believe 2+2=5, politicians could make 2+2=4 illegal, thus we should take that into account when making engineering decisions?
