Lock-in should be broken. CUDA is one of the worst things about this whole ecosystem. Looks like AMD came close to breaking it, but they abandoned developing the translation layer.
It would take away a huge chunk of their advantage, no doubt about it. Let Nvida compete on merit instead of lock-in. Then you can say their advantage lies in being better. But Nvidia is very lock-in oriented, which undermines the claim that they are so much better than everyone.
CUDA IS the merit. They’ve been developing the software stack to make GPU programming accessible for 2 decades now. They’re a software company as much or more than they are a hardware company. Ignoring this fundamentally misunderstands why they’re in the position they’re in now.
It's not the merit - it's the moat (i.e. lock-in) as the linked article states. Merit in this context would mean something you can compare across different GPUs. For CUDA - you can't. I.e. it's a tool to force you to use Nvidia.