The way I understand it, this is useful both in time-critical emergencies (like keeping the grid at 50 Hz when a nuclear plant suddenly goes offline for some random safety reason) and when things are moving a lot slower, like during periods of extreme colds and extreme electicity demand from heating houses.
So, this approach is useful in removing the need for rolling blackouts, at least until demand reaches some very extreme level, but it does require ahead of time investments in shutdown agreements.
Perhaps Texas hasn't been in this extreme situation before, and they simply don't have enough industrial shutdown agreements in place?