graphene based projects are difficult to evaluate... no one knows how risky you go, and additionally you could waste a lot of ressources. which shouldnt be done in blockchain and crypto at all.
I'd agree, and typically this difficulty in evaluation exists across all the blockchains that claim to be "next-gen" or "orders of magnitude better than Bitcoin" in terms of performance - solving the scaling issue first, before encountering it, is always setting yourself up for trouble. I always think it's better for a blockchain to encounter scaling limits and then see how it performs when implementing scaling solutions.
But I don't think the risk should be a deterrent. Clearly if guys like Gavin thought Graphene was viable, then it's worth exploring. But with 2nd-layers like LN already gathering traction, and a seeming next direction to be privacy, I suspect Graphene's not going to be part of Bitcoin anytime soon.