dashboard: restructure repro retesting configuration

Currently we wait ReproRetestPeriod both to start retesting and to
retest every reproducer that's still working. That's not very
convenient.

In some cases (e.g. for linux-next), it's often the case that bugs are
quickly introduced and quickly resolved. Let's adapt to this and retest
the reproducer once shortly after crashes stopped coming and then once
in a potentially long ReproRetestPeriod.

Depending on the results, later we'll use this to also close such bugs
earlier.

0 files changed, 0 insertions, 0 deletions