what you haven't thought about is that when you do refactor the 1000 line file into 10 files, that refactor would necessarily have made each bit slightly simpler (by virtue of it being smaller). Hence, it is likely that bugs don't get introduced as easily, simply because there is less to 'worry' about when writing code that has a smaller cognative load. so may be the metric isn't as wrong as you initially thought.
No, you can literally just take 10 javascript functions and put them in 10 different files or one file with absolutely no modification at all. Why would one get flagged and the other one not?
The point is simply that if you have a unit of code that has X bugs per day this will give a lower score if code that is actually buggier is across multiple shorter files.
