Handle P2P Block Requests Better When Optimistic

[nisdas]

Describe the bug

Currently a node which is optimistically syncing has payload root mismatches when serving block requests. The reason this happens is because the execution node doesn't have all the payloads yet.

Has this worked before in a previous version?

N.A

🔬 Minimal Reproduction

Steps To Reproduce:

1. Sync a node using checkpoint sync
2. Once it is optimistically synced and the execution layer is still syncing the execution state, wait for the node to service block requests.

Error

time="2023-10-23 05:20:02" level=error msg="Could not reconstruct full bellatrix block batch from blinded bodies" error="payload 0x7eed9bb37ee7d9930d0c4085332e69e65dd040245748e0e638c492680b1494a8 and header 0x50b1f36deca789d2bb44e51702565e10f0e16c7224a4d202d87fb1cc8bf64d54 roots do not match" prefix=sync
time="2023-10-23 05:20:02" level=debug msg="Could not handle p2p RPC" error="payload 0x7eed9bb37ee7d9930d0c4085332e69e65dd040245748e0e638c492680b1494a8 and header 0x50b1f36deca789d2bb44e51702565e10f0e16c7224a4d202d87fb1cc8bf64d54 roots do not match" peer=16Uiu2HAkutzT7Lppg1R4L5i2ujaUvqkKVvJ8Znmw4MqMmePEP2WH prefix=sync topic="/eth2/beacon_chain/req/beacon_blocks_by_range/2/ssz_snappy"

### Platform(s)

Linux (x86)

### What version of Prysm are you running? (Which release)

v4.1.0

### Anything else relevant (validator index / public key)?

N.A