Add Deneb builder test & update mock builder

[jimmygchen]

Changes

• Add a Deneb builder flow http test, which revealed a couple of issues (see below), now we have a working happy path \o/. I'll look into adding more testes.
• Fix an issue with builder bid signature mismatch. It turned out that the tree hash changes based on the variable list max size. So I had to use a different type alias for KzgCommitments in BlobsBundle, as the KzgCommitments in BlockBody has a different max size:
  

  lighthouse/consensus/types/src/beacon_block_body.rs

  Lines 12 to 14 in 66be8c2

  	pub type KzgCommitments<T> = VariableList<KzgCommitment, <T as EthSpec>::MaxBlobsPerBlock>;
  	pub type BlockBodyKzgCommitments<T> =
  	VariableList<KzgCommitment, <T as EthSpec>::MaxBlobCommitmentsPerBlock>;

• Add missing deserialization logic for blinded block contents
• Update mock-builder to support Deneb types from updated mev-rs and ethereum-consensus libs
• Update shardingForkTime to cancunTime in local testnet scripts

Additional Info

• I had to make some adjustments when running the deneb builder test in debug mode:
  • Use RUST_MIN_STACK=8388608: 4mb thread stack limit wasnt enough, looks like we really need to get @ethDreamer's eliminate blob clone PR in.
  • Increase get_header timeout: 1 second timeout wasn't enough when running in debug - I think we might even want to override this in tests via env vars, as it's likely that mock builder requires more time to return a response.
• There is a separate branch I use for testing with mock-relay that contains expected withdrawals changes cherry-picked from [Merged by Bors] - Implement expected withdrawals endpoint #4390. I've excluded it from this PR as expected withdrawals endpoint change is still under review.

[jimmygchen]

It looks like doppelganger-protection-test are now failing because of the cancunTime change, I'll revert the change in scripts/tests/genesis.json until we have a compatible el version for this test.

[realbigsean]

Woohoo! I reviewed Mark's blob clone PR so maybe we add that one in first

[realbigsean]

Hopefully this gets merged ethereum/builder-specs#87

It'd be nice not to have to introduce this type. It makes things quite confusing IMO