Wild-ass guess: Myths in software engineering

I have seen many papers justifying how much effort debugging or testing costs by saying something like:

“fixing bugs in software account for 50-70% of maintenance costs”

Just checked this source [1], as far as I see, the guy was making a wild-ass guess for the purposes of making an exciting introduction:

“The rapid expansion of computers into virtually all aspects of society has led to enormous growth in the software-development area. Unfortunately, the increases in complexity and capabilities of these systems have not been matched by an equivalent expansion in software-development technology. Thus, these systems are becoming more expensive and many are error-prone. In an attempt to reduce the number of delivered errors, it is estimated that most companies spend between 50- 80% of their software-development effort on testing. These testing efforts are normally identifiable as unit testing, integration testing, and acceptance testing.”

Is there an actual source to this claim? Most likely not. Just another case of someone not checking their sources and just blinding citing what someone else cited before them.

[1] - Evaluating the Effectiveness of Reliability-Assurance Techniques, Collofello