Evaluating within subject variability with only 2 trials

#1
I am doing some within subject research and calculating asymmetry magnitude is a popular thing to do in my field (asymmetry % = (left side value-right side value)/sum of values*100). The issue is that it is pretty variable in some tasks. So I am looking at calculating asymmetry corrected for variance by subtracting the CV from their respective asymmetry percentages. Another issue is that most often only 2 trials are taken for each test, so I am calculating the mean and standard deviation for each individual with only 2 trials. Due to the time restraints and the overall number of our test subjects, increasing the number of trials is not likely to happen. If you only have 2 trials is the CV the best way to go about this? I’ve thought about using the SEM, but from what I understand, that assumes homoscedasticity and that might also be an issue.

There are several subjects that have negative values because the CV % is larger than the asymmetry %. I am trying to separate meaningful asymmetry from noise (also hoping that it's not all noise), but I’m running out of ideas. Any thoughts? Thanks for your help!