Specific Tier Harnessing user's GPU's & RAM resources

What about a OpenAI economic and and indirectly profit perspective: a specific new separate Plus paid tier (less expensive than the current Plus tier) harnessing user’s GPU’s & RAM & CPU resources that would use all OpenAI’s available models along with OpenAI’s resources to compensate capabilities and performance?

That would probably bring a substantial part of OpenAI’s much promised returns to investors while still delivering to users.