Not random.
I have a scheme and a method, worked out from the beginning. Zooming in at the round 2 pairing, the method did not work out entirely.
The plan was as follows: All second best solos from the first groups in round 1A, that did not directly qualify for round 2, let's say the non-group winners that made round 2 via round 1B, would be paired with a group winner from the other half of the scheme. As it turned out (and I knew in advance that this could happen) there are more solos from one and the same original 1A group that made it to round 2 (groups of death ;-) ). So one of these next best solos followed the method, and landed against a designated group winner (decided by a number in the scheme), but the others had to meet other group winners which had no designated opponents because there were groups with no next best solos going to round 2 via 1B (because they lost in round 1B ). I paired these occasions myself.
Short version: Every group in this 2nd round has a 1A group winner in it. Sometimes the opponent was decided automatically because there was only one 2nd best solo from a 1A group. Sometimes the opponent was decided by me because there were too many 2nd best solos from a 1A group, or none at all.