I think that this comment is spot-on. Google doesn't seem to understand that the infrastructure business is fundamentally a trust business: as elastic and transient as the cloud can feel, when choosing a foundation, you're making decisions that must stand for years. Not only does Google provide infamously Kafka-esque service when they're down or otherwise unavailable, they have a history of capriciously killing services and/or repricing them -- and one need look no further than the user dissatisfaction over GAE to know that GCE will be lucky to survive its infancy. Indeed, GCE is caught in a nasty catch-22: because of Google's track record of killing tertiary services, the market is very much waiting for it to become primary for Google -- but because the market is waiting, it may well never be. This is a self-fulfilling prophesy that is as obvious to cloud decision makers as it appears to be lost on Google.