chili: make basis truncation parameters a required user input?

[stestoll]

chili uses a default basis set truncation independent of the nature of the spin system and independent of the presence or absence of an orientational potential (Sys.Potential) or director ordering (Exp.Ordering).

This choice is fine for some of the common simple X-band nitroxide situations, but can be severely overtruncated in other cases.

This suggests that chili could require the user to always provide some basis set truncation information.

Pros:

• Raises user awareness about importance of basis set choices.
• Prevents chili from simulating unconverged spectra from default basis.

Cons:

• Likely a big hassle for users.
• Users might default to using the same basis in different scripts (copy-paste...), without investigating convergence properly.

The extent of this requirement is unclear. Which of the basis set parameters (orientational basis: LLMK, deltaK, jKmin; spin basis: pSmin, pImax) should be required and which can have reasonably general default values?

This appears to resemble Opt.nKnots for rigid-limit powder simulations. However, there, the choice is much less crucial due to the interpolative spectrum construction.