Performance is not the issue. The issue is when it will be ready, and in a laptop you like, and the answer is "not for a while".

Further, while the RISCV instruction set and architecture are open source, that really doesn't imply that  things such as SMM can't be implemented on them. RISC-V M mode will allow such things in fact. I failed to convince the RISCV community that M-mode code should be part of the kernel, in the hopes of making it impossible to have something like SMM. I failed. 

Don't assume, just because the instruction set is open source, that all the problems go away. They don't. System vendors can still do a lot.