In case you haven't figured it out, measuring things in a fair, effective, and accurate way is hard. One of the problems is that we tend to find measurements that are relatively easy to get and use them, even though we know there are serious issues with them.
One potential workaround to this problem is the idea of using weighted measurements. Simply, put, you give a higher weight to more important measurements to give it more significance.
Let's look at an example - something we can all relate to: grades in school. Let's say a student has 6 classes: science, math, english, history, music, and art. The traditional way to give a grade point average is to assign each letter a number (A=4, B=3, etc.) and then divide by 6. However, let's say that this is a school that has placed a huge emphasis on the humanities (for whatever reason). In that case, having the grades for math, science, music, and art have the same weight as english and history makes the measurement less than ideal.
The solution - make the humanities classes count more. Say English and History are worth 50% and the other 4 classes are worth 50%. It makes those humanities classes "worth" more in the grade point average than the non-humanities courses.
The challenge is how to align the weights? Is 50% for humanities too much? Not enough? Does it create a system where someone can game the system by taking the easiest humanities class to jack up their grade point?
There is no right answer - sometimes measurement is part science, part art. However, it is a way to start to make things that are important more valuable, thereby making your measurements more accurate and effective.