some entries protein cal/g off by much more than FAQ suggests
I know cal/g aren't exactly 4/4/9 as described in prior threads and in the FAQ which says it could be "3.8, 4.1, 4.3" instead of 4, but I'm shocked at the degree of difference for some foods. I'm seeing 2.4 and even as low as 0.91 cal/g protein for some foods (butternut squash & sorghum dry grain, respectively). Consider the latter (Food #456569, Data Source: NCCDB). 10 cups (6317 cal) says 186 cal (3%) from protein based on 204g of protein = 0.91cal/g. Mroe than 4x less than the rule-of-thumb 4cal/g. Is this a bug in this entry despite being from the best DB CronOMeter uses?
At the very least, I think the FAQ ought to be edited to note that the variability isn't as small as the range 3.8-4.3 suggests. Is there a paper or website somewhere that lists foods that have cal/g of protein, carb, or fat significantly different from 4/4/4?
I'll just end by noting the importance of this issue: Protein is in many ways the most important macronutrient to try to control intake of with some precision if one wants to optimize long-term health (based on mTOR signaling & lots of other longevity science) and many targets are expressed as % of calories while many others are expressed as g/kg of weight per day. Having g/cal so noisy makes it tricky to control both of these at the same time.