Introduce checkpoint itest

[Kubuxu]

Was attempted in #12659
Hits a weird bug in the mining loop.

[rvagg]

@Kubuxu was the itest in there flaky? I see green ticks on the Checks for that PR and the itest seems reasonable (although maybe the last WaitTillChain could have waited for a couple of additional epochs)

[Kubuxu]

Yes, the itest in that PR is flaky, and both Steb and I spent 10+ hours trying to unflake it.

[BigLep]

@Kubuxu : what's the verdict on this? What test coverage are we missing not having this? From an "insist on high standards", should we get to the bottom of this?

[BigLep]

2025-01-24 conversation: we believe we should have this for nv25 because it isn't handled by passive testing because in passive testing none of finalization is consensus-impacting.