Coumarin is believed to be liver-toxic, but that’s not the case in humans. Some small proportion of humans may be at an increased risk for toxicity if they have polymorphisms that slow CYP2A6 activity, but that risk increase is minor. Enjoy your cinnamon.
The data supporting protein restriction for longevity is not strong. Mice are not appropriate models for extrapolation to humans due to differences in lifespan and metabolism; data in monkeys does not support a benefit of energy restriction on lifespan; there are unknown influences of genetics.
The weak longevity data we have must be contrasted against the far stronger data showing detriments of protein restriction with aging, like sarcopenia and frailty, reduced quality of life, and premature death.
Improving riboflavin status (via supplementation) may make MTHFR work like it should, since the 677C→T mutation simply reduces its ability to bind with its riboflavin-dependent cofactor (FAD).
It’s literally addressing the cause (making MTHFR work like normal) rather than the symptom (supplementing with 5-methyl THF because you make less of it).