Honestly, I'm still struggling with Bitcoin soft fork proposals. I believe we will end up with full introspection: there are too many things people want to build which require it.
But most current proposals are workarounds for current limitations, which will become vestigial when/if we actually fix things. They may be simply unused, or worse, not quite useful. And it's hard to know: if we had restored script and introspection, we could see what people build and then go "ah, this opcode would make this more efficient!", but without that we are guessing.
So I really have to figure out if mevil is real. Serious people have concerns, esp nostr:nprofile1qqsr6tj32zrfn7v0pu4aheaytdnnc6rluepq73ndc2tdjzus34gat9qpz4mhxue69uhhyetvv9ujuerpd46hxtnfduhswulwwv, so they need serious consideration. If I can convince myself it is either not an issue or independent of script power, then I can reasonably purpose what Bitcoin would look like with maximal expressive power.
After that, I can look *backwards* and see if any subsets of that power make sense as stepping stones. I initially thought CTV (well, a more straightforward variant) made sense, as a common case, but brief discussions with Jonas Nick have me questioning whether it actually is still useful with full introspection (or, more clearly, what the right form would be).
As an aside: I think sponsors (done optimally) are necessary for any Bitcoin high-fee future. Feels like a side-quest though!
Sorry I don't have answers. This stuff is *not* simple, the details are critical, and some of our best minds from previous eras are absent :(
