1) The forms are no longer encapsulated and self sufficient
2) You have to pass object references around in order to address the objects
   inside the form and this can cause problems if those references are not 
   cleaned up well.
3) You lose most of the benefits of object orientation
4) You have a truck load of prgs, or one big one with a truck load of procs in 
   it.
5) You cannot see the code in the form designer or class designer
6) The class browser cannot show you the code
7) Form classes cannot be reused without also copying the supporting procs
   thus removing the advantage of subclassing versus cut and paste 
   programming.
8) Every form must have the specific proc calls written into it, IOW using a 
   form class with predefiend behaviors is not possible. Even deciding
   to change to using classes will mean essentially a complete rewrite.