We are not saying the number is wrong. We are saying that a field which repeats a single unreplicated figure thousands of times, frequently misattributing it, has not built an evidence base. It has built a habit.
What a real evidence base looks like
Publish the method before the finding. State the base and the period on the face of every chart. Say which observations are controlled, which are before and after, and which are simply asserted.
Then let people check. A number that survives being checked is worth more than ten that have never been examined.
Why this matters commercially
Boards are being asked to commit capital on the strength of figures nobody can trace. The first time one of those figures is publicly dismantled, every adjacent figure loses standing with it.
That is a foreseeable event and it is worth being on the right side of.