Failing tests in Measurement ID analysis.

[weinbe58]

Currently there are still many failing tests in the measurement ID analysis after merging PR #557 for a similar reason why we needed to update the address analysis for the refactor of the qubit dialect to use qubit.new. The problem is generally that when going to a standard library the analysis doesn't properly go into the call-stack to evaluate the standard library functions being used in side the entry point.

I think the best way to handle this is to merge the qubit address and the measurement address analysis into a single analysis pass like what was proposed in issue #517.

[david-pl]

As discussed, in order not to block the release of v0.8, we will in a first version aggressively inline things so the tests will work again.