Somehow... with the Google product price changes (price hikes!) in maps and app engine, I am not sure if they have enough credibility left to pull this off.
I believe there was a decision that each Google product should be priced at least at cost. So, products that were under-priced had to raise their prices. I'm more comfortable knowing that these products aren't loss leaders.
That said, it was definitely painful, particularly for people that had optimized specifically for the pricing quirks of the old AppEngine pricing model and really got hurt by the new model.
The problem is that you don't know if this cloud offering is priced at-cost, or it has promotional pricing to kill the competition (AWS) and a pending increase is coming once they get scale.
People assume cloud servers are a commodity, but at $150'000+ company cost per software engineer, there's a lot of API and development that is happening to support the initial roll-out of cloud infrastructure, and we're not yet at the stage where such development costs are negligible compared to the total costs (Amazon is getting there, Google is just starting out...).
Probably best to stick to design/services that allow relatively painless movement to competitors. The competition alone would likely keep prices down so that movement isn't needed.
