In the world of research and academia it may often be taken for granted that agricultural production estimates are accurate. The same problem occurs when we rely uncritically on poverty estimates provided by national statistics offices.
This Guardian article on System of Rice Intensification raises the question of whether production estimates should always be taken at face value. While I'm in no position to comment on the particular case that is raised, it is crucial for researchers working on projects that depend on agricultural output estimates to make themselves fully aware of the methods used to estimate production volumes and, in turn, the potential issues of reliability and bias. Beyond this, it seems only sensible to seek additional information that might help to corroborate findings.