I guess I'm "the other guy" because I also liked Pocket (originally Read It Later), though I stopped using it a few years before it got killed. I agree that Mozilla buying it was a mistake, though I try to be charitable and hope it was partly because they wanted to improve the privacy and openness of a product that could, as you suggest, be used to harvest user information. An open web is their mission, if you believe it.
I've looked around to see how much Mozilla paid for it and can only find "an undisclosed amount". Where are you getting the $20M number?
One thing I think most people don't understand about Pocket is that it was more than just a bookmark database. When you saved a site, it really saved a snapshot of the entire page so that it was available even if changed or removed later, and pages were cached so they could be read offline. These were the unique "TiVo" features of Read It Later that made it more than bookmarks.html and also required backend support, which is why the "just make it an addon" argument doesn't really hold water. Someone has to host the service and data, and that (plus the existing userbase) is really what Mozilla was buying.
But all that said, it was a boondoggle and was probably just Mozilla trying desperately to find ways to be relevant in "mobile". They could have easily built this functionality themselves at a cost much less than whatever they did pay for Pocket. Eventually shutting the whole thing down was the shitty cherry on top.