Earlier quoted context omitted.
Google have been trying to find something apart from advertising since pretty much day dot. It's a strategy pioneered by Microsoft, I call it "spaghetti cannon". Turn your free cashflow into spaghetti, fire it at a wall, hope something sticks. It's sorta-kinda worked for MS. They have 4 or 5 distinct lines of business, each of which could qualify as an F500 alone. Google don't have that. They have 1, and only 1, line…
I do wonder sometimes what Android could have meant to them if they had kept it more close hold and made and sold their own devices. Sure they got lots of other companies to do the expensive hardware R&D, manufacture and marketing, but in the end, what's a multi-billion dollar market in devices is still just a funnel into line-of-business #1.
If they had attempted to do their own hardware, it would have been a horrendously uphill battle against the phones in the market at the time (Apple had tiny share, Nokia colossal).
Putting out the OS meant that hardware manufacturers could compete with different features whilst still (theoretically) having a common OS that users could mostly navigate around and app developers could write for. Contrast this with Nokia at the time, where S40 etc. worked wildly differently on different phones, with different JSRs supported by the J2ME runtimes on each phone, and where they essentially competed with their own other Nokia phones for specs; Nokia did a massive range of phones that were mostly the same but looked different or had very mild changes - it was too much choice!
This way Google could carry on doing software and leave the hardware to external companies that were used to doing hardware. I still have my TMobile G1 though, for history's sake. Don't use it though - it practically has no RAM left over after boot.