Long enough ago that I do not remember the results exactly, I tested
a number of leather sealants.  I made a frame to tightly hold a sample
of pouch leather and used an accurate flow meter to measure the
leakage.  I think I used an electronic flow meter but not sure if
I had one that long ago.

I tested the following sealants: thinned rubber cement with and
without talc, talc alone, DOW 111 vacuum grease thinned with a solvent
(probably acetone) and beaten egg white.  I allowed the sealants to
dry and tested and then removed the leather pieces and crumpled them
up and flexed them to simulate mechanical action in use.  I then put
them out in hot sun for a few days as a minimal accelerated aging test.

I can not remember how they compared other than that I think rubber
cement was the worst after a few days and that, although egg white
was not a tight a seal (it might have been better with multiple
applications), it did not degrade as much over the few days and
mechanical treatment as did the other sealants.

It would be great if someone were to repeat these tests a little more
"scientifically".  It is not just the absolute level of seal that is
important since few applications require a total air tightness.  What
seems to be more important to me is that the seal does not degrade over
time and also that the sealant does not stiffen the material if it is
used where this is critical.

Ideally a number of pouches should be glued to a common pouch board
that is known to be leak free and the seal should be applied in the
same manner as it would be in rebuilding work.  Testing should be done
shortly after the sealant dries or cures and then again after extended
mechanical operation.

It would also be good to test the stiffness of the pouch before sealing
and also after sealing and then again after a lot of simulated use.
I imaging that relative stiffness could be determined by the amount of
air pressure requires to lift the pouch with a uniform weight applied.
Probably a good idea to check valve response time since as well.

If I had more time, I would love to build a test setup and test
a bunch of materials but that is unlikely to happen this lifetime.
I have a fair amount of experience with electronic pressure and flow
measurement and have been playing with Arduino microcontrollers for
about eight years now.  I can design both simple and complicated
measurement and control systems.  If someone wants to collaborate on
a project to measure leakage systematically I could probably find time
to design the electronic end of the job.

Spencer Chase