These server profiles are based on the vms in your cluster.
Large = The largest VM in the cluster
Small = The smallest VM in the cluster
Medium = The vm size in the middle of the above two
Average = the average size of all the vms on the cluster.
Based on you policy vrops will work out how many of the above will fit on your cluster for capacity management. purposes, I dont like these as they can give false reading if you have a cluster with 2 or 3 vms on it as the figures get messed up. We deploy specific sizes so i have created custom server profile sizes ( 2cpu x 4 GB ram etc) so i and get a more accurate cm count on an empty cluster.
When it says you are over by x that is how many large vm sizes you have over provisioned the cluster by
