Gross Weight

If it is structural issues -- which I would suggest it very rarely is
-- you'll still have a huge safety margin when 5% or 10% over-weight.
You are not going to be getting that close to the 'g' envelope, and
your landings are hopefully not going to be hard enough to be given 10%
of collapsing the gear!

The answer to this is - it depends. I agree - the average flight does
not take you anywhere near the limits of the g-envelope. However,
momentary loads of 3 gees or more are not unheard of when flying in
moderate turbulence. So for a VFR flight under a stratus overcast,
sure, I wouldn't worry. For an IFR flight in cumuliform cloud, with
scattered embedded T-storms, I would reconsider. Note that while the
ultimate design load is 150% of the rated load, there is no requirement
for the structure to withstand the ultimate design load without damage.
Deformation is permissible. Repeated deformation due to excess loads
may be a problem. This all assumes the key structural components were
correctly manufactured in the first place, and have not deteriorated.
With an aging fleet, that may not be all that valid. However, I will
grant you that for a utility category aircraft, this is not an issue
worth considering.

The same issue comes up with regard to landings. Long smooth runway in
daylight and light winds, in a plane I've flown before many times? No
problem. Unfamiliar airplane and short strip with gusty crosswinds? I
think I want all the protection I can have. How tough is the gear,
anyway? When effectively the same gear is being used on an airplane
with a significantly higher gross weight, that tells you something (the
gear has plenty of margin). When you have a max landing weight lower
than the max takeoff weight, that tells you something too (the gear has
no margin - it is maxed out). Just something to think about.

Further, if you look at accident reports where over-weight operation
was a factor, I doubt you'll find many where structural issues came
into play. In fact, I can't recall reading a single one,

See the NTSB references in my reply to cwk.

Michael