Make Seq type less restrictive, based on audit. about rfcs HOT 3 OPEN

I'm going to assign myself to this one.

from rfcs.

I've experimented with a few ideas and I think I'm going to split them into two actual RFCs: the first about tidying up the Seq interface and resolving some inconsistencies between different sequences (e.g. Array.remove vs. String.remove), and the second about sequence views. Since I expect the changes I'll propose to be much more comfortable to use with subtype checking and generic type inference implemented, I'll implement these before jumping back into sequence stuff.

from rfcs.

@Praetonus before implementing subtype checking and generic type inference, you may want to rebase and work from on top of my refer pass branch (ponylang/ponyc#1715) to avoid a lot of merge conflicts later.

from rfcs.