"Paul kgyy" writes:

In level flight, there's no pressure differential in the VSI, so I
don't see how a leak could matter. Doesn't the case pressure in level
flight always match the pressure in the static line?

Well, if the leak is toward the cabin, then that is a source of
slightly lower pressure. Still, gazing at a diagram of how a VSI
works (http://www.ae.su.oz.au/aero/instruments/VSI.html) it seems to
me that a leak between the case and the cabin would result in a
*climb* indication if anything. But maybe that diagram is not
general.

Perhaps the explanation is something like this. When at high
altitude, the pressure inside and outside the canister are both
relatively low. Subtracting two such small pressures may have
exaggerate mechanical limitations/errors (sort of similarly how
adding/subtracting two large floating point numbers magnifies relative
errors).

- FChE