Points of measurement are a challenging issue.
One of the ... beyond annoying ... aspects of our track-everything-individuals-do-and-utilise-it-against-them contemporary information ecology is that it is so painfully difficult to make use of that information for personal advantage.
In the specific case of food intake, it should be reasonably trivial to aggregate purchase information, at grocery stores, restaurants, and online deliveries, and at least arrive at a reasonable baseline of total consumption. Rather than having to fill out a food diary from memory with uncertain measurments, one can rely on grocery and menu receipts directly.
This is more useful for those who live alone or shop for themselves (a large fraction of the population, but far from complete). It's based on the general principle that you tend to eat what you buy. There's some error imposed by food acquired elsewhere (shared at work, school, from friends, etc.), and of tossed food, but what you'll arrive at is over time a pretty accurate record of intake.
I'm surprised that such methods aren't more widely used or reported in both dietary management and research.
My own personal experience has been that I've been most successful in dietary management when 1) I have direct control over shopping and 2) I focus far more on what I eat than how much, though some of the latter applies. If I'm aware that specific foods are deleterious to goals (highly-processed, junk foods, high-caloric / high-sugar liquids, etc), then the most effective control point at minimum decision cost is at the store. If you don't buy crisps, chips, biscuits, fizzy drinks, ice cream, and the like, it's not at the house for you to consume.
I'm well aware that there are circumstances in which this is difficult to arrange, sometimes with friends or roommates, more often with families. I'll only say that clearly expressing terms and boundaries is tremendously useful here.