Update snapshots CI run tests sequentially and now it takes >2h which is too long. Update known good is constantly failing because there is always some network is changing.

We should have two known good file, one for kusama networks and one for polkadot networks and have 2 sets of update snapshots / update known good CI.