op-node: Ensure Sync Target

Verifier wants to sync till the tip of the chain. More formally, verifier wants to reach the unsafe head.

Verifier may advance its unsafe head via multiple ways.
1. Verifier will read the L1 and advance safe heads, and unsafe head will follow.
2. Unsafe head propagated via p2p is appended to safe heads, and finally reaching the tip of the unsafe chain.

When verifier is started, sequencer(L2CL) pushes its latest unsafe payload via P2P. Since it is most recent, there must be a gap between verifier unsafe head and sequencer published signed unsafe payload.

The ReqRes sync(RR Sync) is the way to fill in this gap.

This issue aims to track, how should the verifiers get unsafe blocks/payloads when sequencer is down.

Not a blocker for removal of RR Sync. This behavior is for the handling EL after snap sync, the initial EL Sync run. So we may deprioritize this.

No file or test is named in the issue. Start by tracing ReqRes sync and the initial EL Sync path after snap sync, including how P2P unsafe payloads advance safe and unsafe heads. Done means a verifier can obtain the missing unsafe blocks and reach the unsafe head when the sequencer is down.

Written by the indexing model from the issue text.