Instrumentation Cable with Multiple Signal Drain Wires

Usually you can run all the drain/shield wires in common. Then there's the odd times where n isolated shield is needed on one signal or another. My experience is that the few times an isolated shield helps is with an analog signal shield being capacitively coupled to ground. Then there's also the rare times where a deliberate ground loop couples in the right interference to cancel a power line signal coming in through another path. Noise reduction is sometimes a little weird.

My colleague wants to maintain the Mylar/aluminum shielding from the device into the junction box all the way up to the terminal blocks and then from the other side of the terminal blocks to the end point all while electrically isolating their drain wires. I believe this is too excessive. His rational for not connecting drain wires is that signal noise on one shield will go through the drain wires on other pairs which would induce unnecessary noise. This doesn't seem correct to me.

Does the capacitor turn it into a high pass filter?

Shielding is a black art in signal integrity. A lot of environmental and signal magnitude information must be known that can only be obtained by standing in the environment and having the experience of dealing with the noise. I think that I have the experience but I know nothing about your environment. [Don't bother to try and explain the environment.]

Unless your colleague is actually a subordinate, humor them and run the shielding their way. Noise/interference problems are dominated by the loudest source in a circuit. Most likely the difference between your approach or your colleague will not be the dominant interference/noise source. If they are your subordinate then it is your call to humor, ignore or attempt to humiliate them with what you know.

If your colleague is your peer or superior, ask them if you can make a comparative analysis between the two approaches that includes an FFT analysis of the analog signals. Likely you will find that the difference in background noise will be irrelevant. If they are not irrelevant then either swallow your pride or gently explain that you have a way to improve signal integrity with minimal cost.

I think the common method is to ground the signal shield only at the source end to avoid ground loops.

http://en.wikipedia.org/wiki/Shielded_cable

Keep them separate. Do not assume that no ground loop will be created.