I've investigated using a spread-sheet (MS 'Excel'
to calculate change-wheel combinations.
'
It proved an interesting exercise, showing some threads apparently awkward for the equipment available can be solved with pitch-errors within tolerance or feasible by part-cutting then die-finishing, at least over short distances – more common than long threads. (E.g. studs, pipe-fittings.)
Some unlikely threads even proved numerically possible, though perhaps not to aerospace standard, on my EW 2-1/2" lathe, limited by its 8tpi lead-screw and change-wheels of 25 to 65T X 5T increments.
Using a spread-sheet, including such tools as absolute cell references, allows you readily to model the effect of changing just one variable so as to minimise the errors, and show them singly and cumulatively.
I'm a bit rusty on Excel now, but it may be possible to plot one " side " of the true and calculated threads at a readable scale, as two superimposed traces on a Cartesian graph, to show where the inaccuracy starts to exceed tolerance. NB: this is not a design-drawing, but gives simply a series of sharp triangles on the thread's axial plane.
(I once used a manually-drawn version of such a graph to establish using a metric taper-tap to "pilot-tap" a hefty BSF thread in heavy steel plate with existing but slightly undersize holes!).