Building with newgcc, preparing for code coverage checks

[JeffreyBroll]

A few familiar problems popped up when I ran a fresh build with GCC 16.1.1 and "--coverage -g" so I had to bonk a few things back into place

• some familiar incomplete function prototypes, but now Lgm_SummersDxxThings don't fuss about incompatible types
• some char* vs char*[] type whining
• just went ahead and murdered praxis, which leaves two Lgm_FluxToPsd routines unable to use a LGM_P2F_MAXWELLIAN fit type; I can restore the file and fix it up right but it seemed appropriate to check first (due in comparable parts to the dubious origin and the somehow-out-of-tune-theremin soundtrack implied throughout)