Just a correction: Google's slogan is not "do no evil", but "don't be evil". The first evaluates only the actions of the company, but the second evaluates the company itself.
For one to not be evil, does that entail the exclusion of all actions that can be construed as evil, even if it is perceived that doing so is for a greater good?
I'm going to say yes that not being evil means at no time doing evil, no matter what the expected outcome.
The point is that it's an existential judgment -- that they would seek to have the question "Is Google evil?" answered negatively in consideration of all things, not that they need to avoid any blame or transgression.
I think the "DBE" topic has been rehashed about 10^6 times more than necessary, so I think I'll just leave it at that.