The general idea ("bugs found later tend to cost more to correct") is uncontroversial.
It's also accepted that this is a very difficult thing to study. There are so many subjective factors involved that's it's really hard to quantify the data or the results. But it seems to me that if Software Engineering as a wants to make progress methodically, it needs to throw out the old unsubstantiated assumptions in order to make room for new conclusions with a solid basis.
It's also accepted that this is a very difficult thing to study. There are so many subjective factors involved that's it's really hard to quantify the data or the results. But it seems to me that if Software Engineering as a wants to make progress methodically, it needs to throw out the old unsubstantiated assumptions in order to make room for new conclusions with a solid basis.