I think that's the point, actually. (Unless there's some political issue that I'm missing.) Google can make a simple mistake and the customer has to bear the burden of it. When something like this happens, there's no indication that Google will or even can address the mistake in a timely fashion -- or maybe not ever. Google's basic philosophy of no customer service for free products means there isn't much you can do, other than wait and hope, or raise a stink and hope.

Since it was supposedly reviewed and denied, it wasn't accidental. There may be mistakes of fact made by Google or different reasons than those the target has complained of, but the opacity of the process makes it impossible to determine either of those.

I didn't ascribe anything to malice, though I do miss the "Don't be evil" Google of the early 2000s. It's more an issue of mass automation and the errors that accompany it. This is a company with 93 billion dollars in cash on hand and billions of dollars of profit per quarter that doesn't even offer email support for forced closures of long standing accounts.

An accidental shutdown can still fall under overzealous use of power.

"accidental" launched a nuclear bomb...

You're missing the point entirely; even if it was accidental, would you have access to your accounts? Would your email be a reliable way to reach you or verify your ownership of another account? Would people able to access the content you've posted?

The answer is quite obviously "no," and that doesn't cut it for people need rely on services like these for their everyday workflow. We wouldn't consider "accidental" AWS downtime acceptable, so why will we accept it from Google?

Although not terribly enlightening, we can still ascribe malice to an algorithm as a shorthand for its essential behavior. In the end it's a black box, can you rule out anything?