Very good questions indeed.
What I have in mind, is the following situasjon.
Currently we are building up a set of calculation sheets for structural building designs, which mainly consists of a set of workbooks for the different materials (steel, concrete, timber). This workbooks calculate the design resistance of the different materials.
In addition we are doing the same thing for sections as well, where specific section data needs to be calculated.
On the other hand I've done a workbook which calculates the forces (shear forces, moments) and deflections.
I need to get over a set of specific data from the material side to the calculation side already (stiffness data, modulus of elasticity). But later on I will also need to transfer some calculation results (maximum and minimum moments, deflections) back to the material design to do the verification of the capacity.
Of course I could simply implement the whole bunch of calculation and dimensioning in one workbook, but I think that's not a good idea.
I like the idea of keeping things apart, and just transfer the necessary data.
One principle I'd like to follow is that the calculating worksheet/program itself decides which data is good for further processing, and make that available.