Am I mistaken in my impression that the bulk of innovation actually happens outside of titans, which the titans then buy? So even hamstringing them wouldn't actually stifle innovation?
On the other hand, when it comes to product innovation, it definitely seems like the bulk of it happens, as you said, outside titans. Again picking on Google (sorry), V8, Google Docs, Google Maps, YouTube, and Android were all acquisitions! And though Google acquiring them might certainly have given them large amounts of resources to expand (e.g. subsidizing YouTube), there is definitely the question of stifling innovation and competition. Consider TinEye, who first published reverse image search; later Google just poured engineers onto their own reverse image search platform and copied it.
I think you can see the same dynamic with Facebook copying Snapchat features. It doesn’t seem like a healthy market.
EDIT: V8 might be better characterized as the product of an acquihire, actually. It might not fit with the rest of the list, which were all literally acquisitions of products developed outside Google.
I occasionally need to reverse-look-up images (mostly album covers without names) and tineye, in my personal experience, is vastly better. It finds matches for the exact image, with good context, etc. Google usually does pretty OK at finding similar images, but is less fantastic on matches, and if so those are joined in the noise of other 'close images'.
To get back on topic: google is best at pretending they're the best. in my experience there's better reverse lookups. ddg is less 'fuzzy', but offers a much better experience with !bangs, and because of the non-fuzziness they're quite good at finding things when you know more or less what text you're looking for. OSM is often better at local details, especially in smaller/rural areas.
Point being: google can throw whatever they want out there, and people will use it, and google says it's the best so people believe it. Because it's on their phones, because google suggests it, etc. I think the EU has a good point with their anti-trust/android business.
If Google's product was better, wouldn't that mean it increased competition and innovation?
It also is a distortion in that it discourages investment into technical innovation if you are trying to make money as a non-titan—-why even bother if Google will just come along and copy your product then give it away for free (to drive traffic to Google search)?
There is still the potential of innovation in this case from academic researchers, who might not care about commercialization. But such a situation where individuals cannot hope to compete and where research just feeds into titans would be pretty awful.
I personally find the whole "companies don't innovate, they acquire" argument weak at its core. You hear about the acquisitions obviously, and can see them develop into products.
Sure, there are apparently 5 or 6 acquisitions involved with Google Docs, but the existence of an acquisition doesn't mean that the acquiring company wasn't able to innovate, or whatnot, it may mean that they had complimentary technology, or were working toward different goals, or its simply a buyout of a potential competitor.