I've been working on a theory that they produce awful UIs as a disinformation campaign. Consider a scenario where an unmedicated paranoid dude thinks GOOG is sniffing his wireless, recording whereever he goes on the internet, and sharing it all to the NSA, CIA, DHS, and his local HOA zoning committee. Unfortunately this is all true, except for the HOA thing, but the dude is crazy so the point isn't if he's accidentally correct or not. Lets say he tries to spread the paranoid meme to an angry gmail user. The gmail user is going to respond "heck no, these guys can't operate some incredible elaborate spying program, they can't even write a simple email UI after email UIs have been in continuous development for about four decades". Thus the public impression of the "don't be evil" is maintained.
I think if the Mighty GOOG ever divorces itself from .gov / .mil then we'll see them release sane UIs.
Unfortunately I'm not sure how my theoretical model applies to .com with UI issues like APPL. Perhaps they're a mere manifestation of the Cthulhu Mythos. Or something else... it would hardly be the first time that totally different root causes leads to similar appearing results.
I think if the Mighty GOOG ever divorces itself from .gov / .mil then we'll see them release sane UIs.
Unfortunately I'm not sure how my theoretical model applies to .com with UI issues like APPL. Perhaps they're a mere manifestation of the Cthulhu Mythos. Or something else... it would hardly be the first time that totally different root causes leads to similar appearing results.