Underestimating the importance of and challenges of support is sickeningly common in the "but we're so smart" circles of the world. Apple "solved" this by reducing the need for support by simplifying their offerings with reduced execution paths, one-size fits most, etc. Microsoft "solved" this by charging massive amounts of support fees to the entire world that wanted more than an IVR-based FAQ when they called for support. Beyond this, Microsoft solved the issue by repeating their version of the issue which usually goes something like "what issue?" Google, on the other hand, has virtually no B2C support experience in comparison and true to history's form they completely downplayed it's importance.
My personal view on this based on what I've read so far is that Google completely screwed the pooch on this rollout. It reminds me that no matter how powerful, no matter how deep the pockets (IBM, Oracle, Worldcom), stepping outside of your core competencies as a company is perilous. Doing it with such an enormous net without appropriate levels of ramped-up user acceptance, usability, stability and other real-world testing just sounds like an idea born of a patchouli think-tank in Silicon Valley, imho
