It seems like the ideal approach would be "proving" syntax proposals by implementing them in a preprocessor first and then later trying to standardize the syntax if it gains traction. It sounds like there are technical reasons for why that didn't work for the case of SCSS-like nesting, which is disappointing but understandable... if that's the case I think the right approach is to take these new proposals and "go back to square one" (preprocessor) vs. "go straight to standardizing something new".