It's not even about that. Google's services are delivered over the Internet, so they're reliant on someone else to deliver their traffic (and advertisements). Their corporate strategy playbook in those cases is to either commoditize or disintermediate their market dependencies.

In this case, I think they were looking at Fiber more as a "model" for other overbuilders (similar to what they did with the Nexus phones). If Fiber was as successful as Google had hoped, they'd probably open source their operational model since it would effectively commoditize the last-mile access business and give Google more negotiating power with ISPs (without having to invest a ton of money). It was never a business they wanted to be in long-term.

Obviously, it didn't work out the way they had hoped. My guess is that subscribers didn't flock to them the way they had imagined. The Internet is a big echo chamber on this issue; and of course anyone commenting on an Internet forum is going to be technologically adept enough to care about Gigabit broadband. If their market analysis was done online, I can see how they might get false signals due to a sample set of "frequent Internet users" rather than "people who pay for Internet access". It's a subtle distinction, but actually quite relevant since the people who pay for but don't use Internet access often are the ones who make the whole model work.

When you get out into the real world where telecom service is measured in homes passed and take rates, you realize that those "vocal Internet users" make up maybe 10-20% of the market. ISPs make the bulk of their profits from Grandma Mabel who pays $70/mo for a triple play bundle and uses 100 MB of traffic a month. Households like Grandma Mabel account for probably half of an ISPs customers, and Grandma Mabel is not going to get any additional benefit from Gigabit Internet -- so she's not going to pay any more for it.

The big ISPs haven't offered gigabit residential service until recently because there was simply not enough demand for it to justify the expense. I think Google assumed that was just a negotiating position, and that there was a larger, unserved demand for high-speed residential connections. But either way, faster Internet is better for Google's core businesses (YouTube, search, etc.) When you own 70%+ of your market like Google does, you become more concerned with growing the market overall than with growing your share. And considering broadband infrastructure is a capital investment that holds its value pretty well, they honestly had very little to lose. But it's telling that they never invested more than a pittance into the Fiber project as a whole.