op-e2e: fix ActiveSequencerCheckDuration

When `ActiveSequencerCheckDuration` was introduced it was left set to 0 in programmatic configs, and it seems like this causes instant checking / hot-loops in op-e2e.
The `Check()` method should also ensure it's non-zero, so it cannot have the wrong value anymore.

This applies to both the batcher, proposer, and potentially other services.

Start by locating the programmatic configurations and Check() methods for ActiveSequencerCheckDuration in the op-e2e batcher and proposer, then inspect whether other services use the same setting. Run the relevant op-e2e checks or tests; done means the duration cannot remain zero and the services no longer enter instant-checking hot-loops.

Written by the indexing model from the issue text.