Here's an example close to my heart (and please note that I'm just a cog and this is not a statement on official strategy -- just my personal observation): In Google Chrome, we decided to do native beautiful polished UI on every platform - Windows, OS X, iOS, Android -- even Linux!! This is a very expensive way to build a piece of client-side software, and some might say the result is only a few percent better than what could be achieved using a cross-platform UI toolkit. But it's a very expensive couple percent to duplicate.
Same thing with multiprocess. Only makes the software a little bit better, but again, extremely difficult to replicate.
If you look closely, I believe you will see this strategy being used by the leaders of many industries that would at first glance appear to have low barriers to entry.
Btw, the reasons I want to move away from chrome are the "sign in to chrome" movement, I'm assuming this is an end-run around the movement to block internet tracking, and the lack of a strong extension model.
Key supposition there. Because of the economic profits enabled by monopoly, corporations and individuals will disproportionately attempt to lodge themselves into niches where there are economies of scale, network effects, and legal/regulatory hurdles. Or lobby to legislate them, or create products that generate network effects and economies of scale.
Of course, not all markets are susceptible to any of those threats. But as a result they're open to competition, which quickly eliminates economic profit and makes companies occupying those niches much poorer and much less powerful.