Finish and productionize L1BoundSum implementation

[tgeoghegan]

#1292 adds an implementation of the L1BoundSum FLP and Prio3 variant. This implements a the spec and test vectors from the tip of this PR branch on martinthomson/prio-l1-bound-sum. To finish implementing this, we would:

• keep up with any changes to the draft that get published to the IETF Datatracker
• adopt the VDAF's code point once one is assigned (updates to Prio3L1BoundSum #1328)
• remove the experimental feature gate

[divergentdave]

I proposed an algorithm identifier for this in ietf-wg-ppm/draft-ietf-ppm-l1-bound-sum#13, and circulated it on the PPM mailing list.

[tgeoghegan]

#1328 updates the algorithm ID.