> Am I mistaken in thinking we would rather have all the garbage in op return vs elsewhere

I don't agree with that. I don't want them to use bitcoin's blockchain to store their spam at all. If the choice *was* "op_return or elsewhere," I would choose "elsewhere," and by it I would mean "somewhere that isn't bitcoin's blockchain." But in fact spammers have another choice: don't post spam anywhere at all. That is the outcome I prefer.

> If filters drive spammers to pollute other parts of the chain can you actually claim that filters have worked?

This is the fallacy of false dichotomy. You act as if spammers have only two choices: spam op_returns or spam more harmful places. But they have another choice: don't post spam on bitcoin at all.

> If people want larger and cheap data storage currently it is my understanding that they wrap data in taproot scripts because the size limits are much larger and the witness discount makes it much cheaper

For values larger than ~153 bytes, yes. But for values lower than that, inscriptions cost more because they require two transactions: the commit tx and the reveal tx. Whereas op_returns only require 1 transaction.