Weight Growth

I found it interesting, though maybe not totally surprising, to read
in a recent AvLeak that the Capsule was something like 3,000 lbs over
weight targets, and that the booster was near it's maximum thrust as
well, which was leading to all sorta of tricks and changes to cut the
capsule weight down.

One was to re-do the entire interstage (sorry, don't have the article
near me to put down the right terms...) thrust structure which holds
the OMS and solar panels and other equipment in an effort to reduce
weight.

To me, it sounded like they were doing a ****-poor job of initial
estimation, and hadn't budgetted for any growth outside their
control.

Which seems ludicrious since the Apollo program had exactly the same
problems and von Braun quietly upped the numbers of the apollo
launchers to address this exact issue.

Is the problem today that they're talking too closely to each other
early on? Or that they're believing each other too much and not
padding the initial numbers enough? Basically, were the capsule
people saying "Yeah, we can do it all with a mass budget of 10,000lbs max"
and the launcher people said "Yeah, we can give you 10,000lbs max" and
they both started AT 10,000 (number grabbed out of thin air)?

Should the launcher people had said "Sure you want 10,000 max, we'll
build for 15,000 since we know you'll go over." Or "Sure you'll get
10,000 no problem. Quickly guys, let's plan for 15,000 but not tell
anyone..."

Or how do you budget properly for weight growth? I read about jet
engines for the A380 and B787 which are supposed to be comming in
around 80,000lbs each, but that GE/Rolls/Pratt have actually run them
at 86,000 already, even though they'll be rated lower.

How hard would it be to just build in a margin and hope to never use
it? Or is the design so damm tight, and the perception of "If we
don't use every last pound of thrust available to launch every last
ounce of mass we can use, we're wasting money?"

I dunno... I just don't have a good feeling about CEV and Ares I at
all.

John