fix(experimental elaborator): Avoid defining globals twice

[jfecher]

Description

Problem*

Fixes the "Duplicate definitions of found" error by removing resolve_local_globals.

Summary*

Additional Context

We don't need to re-add globals to scope anymore via resolve_local_globals since we're not creating a new Elaborator for each function like we do for NameResolvers. The flip side of this is that I think we'll need to change how globals are stored eventually so that they're not all always visible. I think how it is now may violate imports but I'll leave that for a later fix to keep these PRs small. (and wait until I have a test case that shows it).

Down to 61 errors after this PR. The errors are:

• "Duplicate definitions of <generic> found" - looks to be only one case where there is a generic directly on a trait method. fn hash<H>(self, state: &mut H) where H: Hasher;
• "Expression type is ambiguous" - from the two turbofish cases that haven't been merged in this branch yet
• "Expected type &mut _, found type H" - issues with auto-deref and trait methods possibly
• "No matching impl found for ..." - trait solver issues

Documentation*

Check one:

• No documentation needed.
• Documentation included in this PR.
• [For Experimental Features] Documentation to be submitted in a separate PR.

PR Checklist*

• I have tested the changes locally.
• I have formatted the changes with Prettier and/or cargo fmt on default settings.