Aggregate transcript from a subset of participants

[cygnusv]

As part of the efforts to add some fault-tolerance to our DKGs ( nucypher/nucypher#3425 ) as well as for Refresh & Recovery, there is the need to produce aggregate transcripts that don't include all participants in transcripts, but only a subset.

For example, during the proposed fault-tolerant DKG rituals, at the end of phase 1 there may be more than N transcripts, each containing N+B elements, but the aggregate transcript should only contain N elements.

[piotr-roslaniec]

Probably requires testing only

[piotr-roslaniec]

Closed by #185