I'm less qualified to opine on Google than most of the people here, but in hindsight, what Google products remind me of is the way that black walnut trees slowly poison the soil so that the seedlings from other species of tree cannot grow nearby. The good intentions that poured energy into all the 20% products are no longer the point. Somewhere along the way, someone figured out how to use them strategically. The free products are good, good enough to use, until you realize that there is no path for continued growth or investment of resources, and run into seemingly arbitrary disappointments and limitations; it as if at some point, someone stopped the projects from adding cool utility to the product, and started making sure that hindered, crippled versions of the feature were offered instead. I experience this most acutely with the languishing "Google My Maps" product. It feels as if the target is not just potential competition, but the imagination and demand of the market itself.
I don't actually know the story of Google Reader and RSS feeds, but I remember how integral RSS feeds were to the golden era of blogging, and how abruptly that era seems to have ended with Google Reader's apparent death. And to me, that has a similar feeling. The idea is that the target is not potential competition wherever it might spring up; the idea is to sap the demand that might nourish competition, to suck the air out of the room, and stifle the imagination of the market itself.
It isn't Google alone who is responsible for this feeling, to be fair. There is watching the growth of the walled garden of Facebook, watching the collapse of the old chat services which allowed independent clients, watching successful startup after successful startup turn new ideas into content for a routine process wherein we see the exact same sheen of gloss on the promises, the same dance steps towards the pirouette, the attempt to pivot gracefully and effortlessly towards monetization in a maneuver that is in fact a mating dance desirous of acquisition.
All of it really sucks. It's not like there's an easy alternative. People like free things, and with computer-based resources there is often so much opportunity to scale the value of a thing that free things can be sustainable; a project can succeed and be useful to thousands of people merely on the basis of the labour that some are willing to commit to to sustain it. Again, I'm less qualified to describe this than most of you are. But that's what open source is like.
It doesn't work with services. Code that runs of different platforms can be replicated/adopted for infinitesimal cost, and the underlying costs of running it are naturally distributed. Services are different. The replication/adoption and the creation of value both involve on a massive rush of the many to the one. That relationship pretty much sums up the whole story. If capital accrued to capital by a square law, attention would accrue to attention by a cube law. In idiosyncratic niches that cannot be satisfied by the mass service, alternatives are actually viable and flourish. But anything that would be beneficial to us all encounters this problem of needing to absorb the real costs of operation while seeming to be as free as possible, or else the users will flit away to a different flower.
There's no good solution to this, but the way in which Google has graciously assumed responsibility for directing our attention does not make it better. All the improvements to search results over time seem to focus attention more and more to what an archetype of user is likely to be satisfied with. I would not be surprised if the energy costs per search had gone down. As many have noted, esoteric results are increasingly invisible.
Anyways, this is what we have done with the new universe of human communication that has opened up in the last few decades, which we imagined we would leverage into new systems of effortless communication and collaboration. And we have, to a lesser extent. Second best or third best. But we've discovered this really intractable problem with the distribution of costs.