Callaway argued that the KPMG approach was needed only if it had committed an error in setting up the fund--and that a change in return estimates wasn't an error.
No one, not even the Soviets themselves, knew exactly how much it cost but it was a large enough figure that the best estimates about the Soviet defense burden have a margin of error of around 2% of GDP.
As students, you should keep in mind that all these numbers are estimates based on samples, which are thus subject to sampling error (as well as other types of errors).
And yet the article does not mention to the reader that they are estimates and therefore any extrapolation of market share has an unknown margin of error is just wrong.