I know it's hard if not impossible to size a system just using square footage and other general characteristics. I've even been told that doing a load calculation (manual J?) is only as good as the data put in, and that unless you have specs about building material, wall insulation, windows, doors, etc, which is normally not readily known in a 30 year old house, the load calc is hit or miss. What I've been told by the latest tech/sales guy is that a 3 ton might be needed to compensate for a duct system that is not truly balanced. The house originally had one system. When the basement was finished back in the 80's, the second (attic) system was added, so that the new system handled the second floor and the original system handled main and basement. As a result, some of the ductwork has been capped, and some added to the original main trunk.
First two reps were Lennox and last two were Carrier. When I asked about Trane, I was told that it was being sold. I'm still trying to decide which to go with, assuming all priced in same, give or take a grand.
Oversized equipment will ALWAYS make any ductwork or load imbalance between rooms or floors worse, not better. More air flow in the same size ductwork also makes it harder to balance the system. A small unit on the same size ductwork, give you freedom to reduce airflow to registers that have too much but using simple dampers in the ductwork.
A larger unit would only be needed if, after correcting a temeprature imblance, the net heat loss increased. Meaning if several rooms were underconditioned and none were over conditioned, after balancing, all room are slightly underconditioned... meaning your short on capacity.