https://archive.is/bVDzB
A small proportion of people in the US are responsible for eating 50 per cent of the beef consumed in the country, which is a major contributor to greenhouse gas emissions
It seems likely to me that they asked that because they felt respondents would be most likely to recall what they'd eaten most recently, and less likely to be able to do so outside of that window.
Not sure what the real problems are extrapolating that that one day is a typical day given that routinely you extrapolate from a small sample, if that sample is representative, to much larger populations with pretty acceptable margins for error.
I guess the question is, was this question asked at a time or in a manner that would lead us to know that the period being asked about was in some way atypical for the population being asked.
The study design was reasonable. The unreasonable part is implying that what the respondents did that day is what they do every other day. Because what respondents do on that day does not necessarily predict what they do on other days.
In other words, it is reasonable to assert "On a given day, 12% of the population consumes half of the beef produced that day". It is not reasonable to imply "12% of the population consumes half of the beef produced in any given time period".