Unblock Gloas submissions through the existing bid-submission wire shape

[0w3n-d]

Issue: #489 (step 5 of 6, reordered ahead of the original step 5 -- see issue update)

What this PR does

helix's ExecutionPayload/SignedBidSubmission (crates/types/src/bid_submission.rs)
is its own bounded-list builder<->relay wire shape, shared unchanged across
Bellatrix-Fulu -- not a mirror of the real per-fork consensus SSZ types. The
ForkName::Gloas arms in hydration.rs/execution_payload.rs currently hard-error,
which blocks Gloas-fork submissions from decoding at all.

• Routes ForkName::Gloas through the same decode path ForkName::Fulu already
  uses, in the 4 places that previously errored.
• Adds ExecutionPayload::to_lighthouse_gloas_payload and
  execution_requests_to_gloas, converting helix's bounded-list submission data
  into the real, progressive-list Gloas consensus types
  (ExecutionPayloadGloas/ExecutionRequestsGloas), for use at the outbound
  bid/envelope boundary.

What this PR deliberately does NOT do

• Does not wire these conversions into getExecutionPayloadBid or
  submitSignedBeaconBlock yet -- that's the next step.
• Does not populate block_access_list (EIP-7928) or builder_deposits/
  builder_exits (EIP-8282); left empty with a TODO(gloas) marker since
  there's no producer path for any of these yet.

Tests

Written before implementation, per this repo's workflow:

• Gloas-fork decode round-trips for all 3 affected hydration paths, mirroring
  the existing Fulu round-trip tests.
• Field-by-field correctness of both conversion functions (transactions/
  withdrawals/deposits/withdrawals/consolidations preserved, new fields
  defaulted, slot_number set correctly).

Reviewer checklist

• CI (lint, unit-test) is green
• This PR matches the step it claims to be (see linked issue)
• Nothing surprising: no unexplained scope creep, no unrelated files
  touched, no obviously-wrong code the tests happen not to cover