Super interesting - thanks so much for this. I'll admit it's never actually crossed my mind where those raw breakdown numbers come from on the back of the packet (outside of kcal which is the Atwater stuff).

I had a bit of a look through and now I understand your point - that determination of protein content is based on:

- assuming all identified nitrogen is in protein

- assuming all identified amino acids are in protein

using the 6.25x conversion factor to scale up the identified nitrogen to be the assumed protein mass. I had a flick through this paper (https://apps.who.int/iris/handle/10665/331206) - the intention of which is to analyse this 6.25x factor and it's applicability of dairy vs soy - and it agrees that there should be variation in how you scale that nitrogen mass up. Whilst it gives slightly lower values for soy than dairy, it's in the region of around 5% so not massive but definitely an interesting point.