Problem: Forecasting schedule is a difficult thing to do. In addition to other methods; methods for computing software schedule include the use of estimation software from your size and effort estimations and use of historical data - which probably includes experience as well.
When you combine these with algorithmic approaches such as COCOMO, do you think it's possible to get a clear estimate?