Refactor executor setup/executor name function

What does this MR do?

• Removes Name() string from executor interface. Only because it would now be dead code.
• Refactors executor setup:
  • No chance of calling Name() on empty executor
  • No need to attempt to defer cleanup when we don't have an executor

Why was this MR needed?

Setup was slightly confusing and potentially error prone. Part of potentially larger refactor to simplify.

What's the best way to test this MR?

Existing tests should cover

What are the relevant issue numbers?

Closes #30846 (closed)