The problems with us-east-1 have been apparent for a long time, many years. Once I started using us-east-1 long ago, and seeing the problems there, I moved everything to us-west-1 and stopped having those problems. EC2 instances were completely unreliable in us-east-1 (we were running hundreds to thousands at a time), not so in us-west-1. The error rates we were seeing were awful in us-east-1.

A negligible cost difference shouldn't matter when your apps are unstable due to the region being problematic.