There were two types of affected nodes. Those that fell behind with syncing, as they most likely couldn’t keep up with updating their blockchain db (starving nodes). And those that were fully synced, but were asked by those starving nodes to provide blockchain updates (most likely, repeatedly the same data, thus the high traffic).
My box was in the second category. The only way to know whether that was the case was to check stale proofs on the pool side. Thus, we don’t know if those that were self-pooling were affected (but rather nothing points that they were not). There were no indication on my box that there was an issue, as all resource usage was minimal.
That attack ended with 10-20% of network space shutdown. As most likely, those affected nodes were small farms, that implies that maybe 60-80% of nodes were actually hurting.
Saying that, the problem is over for now, so if you are fully synced, you should be fine.