It is difficult to argue with the premise,
"It is never safe to assume that computer calculations are correct."
However your next comment seems to jump the gun - at least for me,
"In the case of this bug, the mathematics is trivial, so one can simply
ignore the bug."
Ignoring the bug when attempting to find a numerical result could be
very problematic. The fact that there is a trivial bug points to the
fact that there could easily be other "trivial" bugs, i.e. there could
be times when 1+1=1 etc.. Can this be ignored? Is this trivial? At
least for the 2nd problem the answer should be, "no".
If this is to be expected, and Sen I believe this is what you are
telling me. Then I must forget thinking of Maxima as a program that
will sum correctly simple trig., exp. and other functions together very
nearly without error because this is not what Maxima is - very good.