Of course the problem of fundamentally building on someone else's platform remains. This is why it's a good idea to start at least thinking about how to move your business onto another platform, from day zero. Of course there's a balance between investing time and money in building abstraction layers and evaluating other platforms and doing these different things, versus spending that time and money on immediate growth. But if one invests literally 0 in this situation, they can hardly expect much sympathy if the underlying platform gets yanked out from underneath them and kills their business.
We have a finite amount of life energy to put into our businesses and I’d prefer to give it to something that can generate an annuity.
Does Exxon stop drilling for oil just because it won’t be possible someday?