> They can, in some cases, do services, but the idea that something ever reaches a finished state in which everything is complete doesn't seem to be part of their DNA.

I know what you mean, but I'd actually say the opposite:

Google has a history of dumping a bunch of resources into a first release. Which may be quite good (Google Books), barely a proof of concept (Wave), somewhere in between (Hangouts), or bizarro what (Plus). And then withdrawing all resources, never iterating on it again, never delivering any major improvements.

In fact, this, it seems, is what Glass-purchasers felt for the past year and a half -- So, this is a good first draft, but when's it gonna get better? Answer: Never.

Presumably the projects which start looking like they will make money for Google receive further investment.