Yes you are correct that incomplete lists (especially very short ones) will bias the results. The size of the bias will depend on how big the weights are for listed composers compared to unlisted composers.
If a weighting system like 1 = 100 pts, 2 = 99 pts, ... , unlisted = 0 pts is used, and only ten composers are listed, then the other 90 composers that theoretically should have been on the list will be biased downward (especially the ones that were barely left off, which will lose nearly 90 points).
One way to reduce bias is to bring down the size of the largest weights closer to the smallest weights (maybe even give a positive score to every single unlisted composer), but the cost of this is that people's rankings will mean less. Since we don't have that many voters, the end result is really just a matter of the weighting system chosen.