log engine/builder API decisionmaking

[tersec]

If this PR is to be useful, it's not clear to me whether it has as much value at debug as info.

The general goal here is to clarify for for Nimbus end-users and for developers if/when Nimbus is properly comparing engine/builder bids and when it's inadvertently selecting one because the other didn't work. It furthermore shows even when selecting one, if both arrived, the value of the other bid as well.

I don't view this as critical functionality, but might be nice to have, and came up during some discussions/apropos concerns about what happened when Nimbus sometimes did not use a configured builder AP relay but fell back to the engine API

General reasons this can occur are relays being down, or more commonly, returning HTTP 204, indicating nothing interesting to suggest for that slot. It can also happen that they do suggest some block (header), but it's not particularly valuable.

The latter becomes potentially more interesting as the local block value boost is increased, because usually in a direct comparison, the builders and relays will create higher-value blocks than local EL clients, but might not if they have to be 15%+ higher-value.

[github-actions]

Unit Test Results

         9 files  ±0    1 098 suites  ±0   27m 17s ⏱️ -29s
  3 965 tests ±0    3 618 ✔️ ±0  347 💤 ±0  0 ❌ ±0 
16 090 runs  ±0  15 692 ✔️ ±0  398 💤 ±0  0 ❌ ±0 

Results for commit 6bc463b. ± Comparison against base commit df902fd.

♻️ This comment has been updated with latest results.

[tersec]

Of note, Lighthouse has something like this: sigp/lighthouse#4352