Fix tuist generate performance regression of ModuleMapMapper

[adellibovi]

Short description 📝

There was a regression of performance introduced with the new ModuleMapMapper. This PR tries to fix it.
Pre ModuleMapMapper I had a dirty tuist generate of <5s, with it I could reach ~18s. This change brings it back to ~5s.
There are two commits:

• one that introduces the use of dictionary instead of first
• another one caches any processed target, before we were excluding the ones that did not produce any dependenciesModuleMaps

Checklist ✅

• The code architecture and patterns are consistent with the rest of the codebase.
• The changes have been tested following the documented guidelines.
• The CHANGELOG.md has been updated to reflect the changes. In case of a breaking change, it's been flagged as such.
• In case the PR introduces changes that affect users, the documentation has been updated.