I've had professors do this to me. I don't understand what's wrong with this. This is a common thing I even do when grading programming assignments. When I notice a stylistic choice that is outside of the norm for the students I can usually tell exactly who the work came from. Every semester I catch at least one group of cheaters (This semester I caught two!) and they admit to it after I confront them with it. "Hence…
What you did makes sense. This is the standard two tier laboratory testing process . The first screening test is cheap with some false positives. The second confirmation test is more troublesome but more accurate. The professor did the first and evidently only the first.
If the student does nothing then I'd assume they did copy.
I'm sure people will be mad about these assumptions but from my experience--and from what I've seen my colleges fall into after wrongly accusing someone of cheating--I'd definitely say this is an appropriate average model. I'd also say it's lived to fruition in this case. The author got mad and wrote a fantastic expose on this situation. If I was the professor I'd correct what was obviously my mistake at this point.