Everyone wants to know how a similar product performed. Very few benchmark sets can answer that honestly, because most of them record what launched and almost none record what was believed beforehand.
Without the prior belief you cannot tell whether a launch succeeded because the team read the market correctly or because they got lucky in a way that will not repeat. Both look identical in the outcome column.
This is why recording the prediction matters as much as recording the result. What did we think would happen, what did we change, what actually happened. Three columns instead of one, and only that version teaches you anything.