I enjoy drawing my schematics on the whiteboard and snapping a quick picture. This is the schematic for transient voltage suppression going into the USB 6008:
Started wiring it up:
The diodes are heavy duty:
The diode’s leads are too thick for the USB 6008 Prototyping board:
I can still solder these diodes on, but it will get crowded on this board:
So I will also use standard prototyping board:
Prometheus Fusion Perfection 
Sorry about that. You could try inserting a smaller solid stiff wire through the prototyping board and solder tack the protection diodes to the smaller wire. A little kludgey but it will work.
That doesn’t get around the crowding unless you mount the diodes vertically.
Putting the diodes on a separate board has one caveat; make sure the lines to be protected go past the diodes before the usb box. That is, don’t have the diodes dangling off to the side; the bad voltage impulses must hit the diodes _first_ and then, after being clamped, allowed on to the signal acquisition unit.
Oh yes, I should have mentioned; been very explicit.
The common side of the diodes should be heavily bussed together and connected to the computer frame via some of the heavy braid I saw you use.
Actually standing the diodes off of a bare copper plate (as in a raw pcb material) and connecting the plate with the braid would be even better. Since the errors aren’t destructive you might as well try what you showed. This is just to tell you that there are better (and better, and better….) (and harder and harder …) ways to do things.
Something I was ashamed to mention before; but it’s been 55 years so….
I built a Jacob’s Ladder from coat hangers and partially broken Neon sign transformer. Since you have a HV supply maybe you could build one. A great EMI generator! Unfortunately the same problem with interfering with the neighbours electronics exists.
The only thing wrong with you setup is that the leads to the USB device would do better if they were twisted; all three together.
I keep thinking the making a Balun would be better; that is twisting the wires and running the whole twist through and around a toroid. They work great up to about 100Mhz; it gets harder above that.
The wikipedia page: http://en.wikipedia.org/wiki/Balun
is fairly good, although missing several points I consider important. Here I am talking about the “transmission line” application as illustrated in the first picture. The wound windings (sorry about that) can be; transmission lines, coax (as shown), twisted pair, or shielded twisted pair. The idea is to put an obstacle (inductance) in the way of common mode currents and force all currents to be differential. The problem with twisted pair and common transmission line usage is the capacitive coupling around the inductance. Nonetheless you should be able to get 20db or more common mode rejection up to 100-500 Mhz (using HF beads rated in that range).
Of course if I were you I would go ahead with a test.
I think your camera, and usage, is neat. I wonder if we realize how far in future we are. People picture the technological revolution as wave; I disagree. It’s more a fractal pattern with surprises at every turn. Using the same reasoning as Mandelbrot.
You might want to try using a diode with a higher peake inverse voltage. I had been losing rectifiers on an elevator in a part of town.known for poor quality power and a stinger leg. I ended up using diodes from a microwave…. never lost another rectifier.