feature request: option on double precision position values

Yes, it is possible today, and the instruction tobspr gave is correct: add --override STDFLOAT_DOUBLE=1 to the makepanda command-line when compiling Panda3D from source.