I represent "a big city hostel" here (in fact three) so maybe I´m slightly biased.
Try to look at it from the customer´s perspective: an average rating based on very few or very old ratings isn´t worth too much. So it is arguable not to publish it at all. I agree that´s a disadvantage for small seasonal hostels, but then Hostelworld should be understood as a site for travellers, not for hostels.
Our average ratings are based on around 1000 ratings for each hostel, so the figure is reliable.
As for the top ten: as far as I know, the minimum number of ratings used to be 50 ratings per month - so, reducing that to 20 is already a big concession for the small hostels. Compiling lists of "best hostels" based on 5 or 10 ratings would hardly be more than a bad joke. In my humble opinion.
I also had a ongoing disagreement with Hostelworld about the possibility to make comments "inactive" (the euphemism alone outraged me). I am very happy that this can´t be done any more. It is a big advantage for the travellers - now they can make a sound decision about where (and where not) to stay (publishing half-truths is worse than lying). This is also a big advantage for good hostels.