I don't know about the numbers exactly, I'm not a scientist! I did buy my 5dii slightly used (shutter count under 5000), so it's always possible the filter has been changed without my knowledge! However I highly doubt that! My test was simply to point a 940nm IR led at it, which it clearly recorded as pinkish white light with red in the middle! Many cameras can actually see IR in the 900 range. I agree with your explanation about the violet in the OP's photo, that's what it seems to be. I'm surprised OP didn't see it, that's exactly how I've seen yellow (tungsten) and blacklight violet/UV mix, I instantly recognized it! The yellow shadows in the deep purple (the object blocking the blacklight but not the tungsten) is a dead giveaway! The only difference to viewing it with your bare eyes is that the tonality is poorer
It would depend on how long you exposed for. Pointing an IR led directly at the sensor, I assume from a relatively short distance, is quite a bit different than picking up IR and UV along with whatever other light is reaching your sensor during any normal exposures. Most modern digital camera sensors use a standard silicon substrate, and silicon is naturally sensitive to a spectrum ranging from ultraviolet (maybe ~250nm) through deep infrared (up to 5000nm). The peak sensitivity range is from maybe 300nm through maybe 2000nm, beyond that the natural sensitivity levels for both deep UV and deep IR are relatively low.
It also depends on the exact characteristics of the IR cutoff filter. As I mentioned before, most DSLRs don't use high quality UV/IR filters that have high transparency to desired wavelengths and a strong and total cutoff to undesired wavelengths. They have a more gradual curve to them...filtration usually starts in the deep reds of visible light, and tapers off into near infrared. It's possible the 5D II has a rather long falloff period that allows more IR through.
It was from a close distance yes, but after reading this I did another test: I asked someone to hold the IR source (a remote controller with a 940nm led, the wavelength of the led was verified) a little further away at a distance of 3.5 meters. The light seemed a little weaker but was still perfectly visible and had both the white and the pinkish component. Checked in LV at 10x magnification using sigma 35/1.4 at f/1.4, 1/50s shutter speed. Inside lighting illuminating the remote controller was very yellow CFL (around 2800K), WB was adjusted to neutralize the color temperature.
I'd offer that 3.5 meters isn't all that far for a concentrated IR emitter like a TV remote. Those things emit a pretty powerful signal, even though we can't see it. They have to combat the ambient temperatures in peoples homes, the energy emitted by direct sunlight hitting the receivers (well, in some cases...one of my IR remotes still works when the receiver is bathed in direct sunlight, the others tend to be sketchy), etc. In terms of infrared light, those remote controls send out a pretty "bright" beam...kind of like a bright visible light of a handheld search light vs. the dimmer beam of a flashlight.
You also have to wonder if your little remote control is just emitting IR, or whether it is emitting some visible light as well. I know that I can see a faint yellow light emitted from the IR LED in one of my TV remotes if I look at it while it's emitting. I think I once had a remote that emitted bright red light as well as IR, as it doubled as the indicator light telling the user that the remote was actually indeed sending a signal.