The bfloat printing issue was addressed years ago.
How do you print a float-like number?
A: If a format string is provided, e.g. CL ~10,4,F or some such thing,
use it. I don't know if anyone does this though.
B: If we must rely on some default, e.g. CL "g" format, I think, there
are
still some possible parameters.
1. fpprintprec.. maximum number of digits to print in the
"fraction" part. If 0, use as many as justified by internal precision.
(commercial macsyma calls this bfprint_precision)
2. fpprintzero
Do you print more than one trailing zero? e.g. 3.0 vs.
3.0000
Maybe this is not useful.
Since maxima usually deals with expressions that are more complicated than
floats, maybe we should implement/document locally binding formats, sort of
like
block ([default_float_print_format:"~10,4,G"],
display(<some expression>);