I'll see all of your assumptions listed and raise you at least half a dozen more.
The 'Garbage In, Garbage Out' thing is (has always been) very true although I've seen it rephrased as 'Warning: The accuracy of this software will depend entirely on the quality of your assumptions.'
The Ballistic Coefficient doesn't vary with velocity (or temperature or pressure or humidity or moon-phase or a plethora of other things)...
BC = Sd/FF
Where BC = Ballistic Coefficient
Sd = Sectional Density = Weight (Lb) / Diameter² (inch²) and
FF = Form Factor = observed Cd / Reference Cd at any particular velocity.
... note that, for any particular projectile, Sd is constant and FF is also constant if (and only if) the correct reference drag law is used. Obviously, if Sd and FF are both constant then BC must be constant too. As Brian mentioned elsewhere in this thread, the GA drag law is probably the best fit for round-nosed, diabolo JSB Exact pellets and their close clones.
If you see curves of BC varying with velocity (common on some American websites the authors of which should really know better) then either the drag law has been badly chosen or, more likely, a (easy to use but wildly inaccurate) constant Cd model has been used.
*** See the earlier remarks re. Quality of assumptions ...
Good luck with that (unless you've got access to a wind-tunnel and/or a Doppler Radar set). I'm told that Chairgun's developer took 8 months and well over 2000 data points to generate the GA curve. Not a simple task.
As above. Temperature dependence would indicate that your drag law is incorrect.
However, you can easily use Chairgun's environmental variables to match the the current atmospherics to re-calibrate the predicted trajectory to reality.
*** See the earlier remarks re. Quality of assumptions ...
George