refactor(consensus): Refactor code in NiDkg to prepare VetKD implementation

[Sawchord]

This PR makes changes to DKG in order to prepare the implementation of CON-1413.
In particular, it makes two changes:

1. Summary::current_transcript now returns an Option, rather than panicking.
   With the introduction of the NiDkgTag::HighThresholdForKey(_), we can no longer enumerate all variants, and we should not assume that the Summary contains all the tags we assume. In fact, when generating a vetkey, we will end up in situations where next_transcripts contains tags that current_transctips doesn't have. This PR simply adds unwraps at all call sites, since the code should not panic, if called with NiDkgTag::LowThreshold or NiDkgTag::HighThreshold.
2. Reimplement into_next_transcripts and move it into dkg crate
   The current implementation is slightly wrong, since it enumerates the tags from the current_transcripts, which makes the assumption that the tags of next_transcripts is a subset of the tags of current_transcripts. This is not correct. When we add a new vetkey, we generate a Config for it at put it in theSummary. This will generate a NiDkgTag::HighThresholdForKey(_) transcript for it and put it into next_transcripts of the next block. So this transcript would never become a current_transcript with the old code. The new code simply copies next_transcripts over to the new summary as current_transcripts and then looks for old current_trascripts whose tags are missing and copies them over too.
   Also the code was moved from ic-types into the dkg crate, since it was only used there, and this way we can add some logging.