Since I will be using this camera in many low-light situations, this will probably become a major issue for me.
Low light is a very relative term. Just what do you mean by low light?
For example, as I just noted, in my previous post, with the meter on CWA I see a stable meter in a room lit by 2 24" LCDs monitors and nothing more (meter reading 8s f/2.8 ISO 100, est EV0).
If you want to frame your expectations, look though the images you have, anything darker than about EV1* (meters 4s f/2.8 ISO 100 or equivalent), is below the design range of the meter and may not meter correctly. Anything below EV -4* (2m f/2.8 ISO 100 or equivalent) is below the meter's hard sensitivity floor. As it stands*, anything between EV -4 and EV 1 have the potential to be thrown off by turning the backlight on. There are of course caveats to this, the actual cases are subtly more complicated but that should put you in the ballpark.
My biggest concern is that Canon Canada has not released any press release whatsoever, and the camera stores still claim (as of today) that this is a rumor and that no real hardware problem exists. I have also not heard of any "new shipments" with design changes coming in.
Keep in mind Canon is a big company, and the regional Canon divisions aren't all in complete sync with each other. Things take unequal amounts of time to work their way though the bureaucracy in different places. There may also be various legal implications with posting something like that in different places, which likely means each region has to run the service notice though legal and that doesn't happen instantly either.
From the various websites covering this, the issue goes well beyond the LCD screen (as the US website describes) to the top display and even the viewfinder, so this is pretty substantial.
I would urge you to be very careful what you consider valid information while reading about this. The barrier to entry for publishing something on the Internet is incredibly low and there are no safeguards to stop the blind from leading the blind so to speak.
The behavior seen in the CameraTown video, where the meter changes as you move your eye away from the viewfinder, is typical and expected behavior in an SLR where the metering sensor is in the viewfinder—read all of them. Some may be less susceptible to light intrusion because they have a smaller viewfinder, a viewfinder that is more recessed, or there are other differences in the internal design (different reflectivity of the focus screen for example). But make no mistake, under the right conditions, and those aren't just "with the sun at your back", removing your eye from the viewfinder will change the meter reading.
I was planning on purchasing this camera this week (prior to my upcoming vacation) but as I will ultimately spend upwards of $6000 on this (kit, accessories, etc), I am pretty concerned about buying this if it is defective? What about all of your thoughts on this? For those of you have who have yet to purchase, has this issue now put you on the fence?
Well it's probably pretty clear that I already own a 5D3, and I'm using my 5D3 with little concern with the reported issue as it doesn't actually seem to be a problem in practice. My 5D3 meters virtually identically to my 1D3 in any situation where I've had the two side by side. Worst case scenario, if Canon determines that they must fix something, there will be a recall and a fix will be made for free, you'll just have the hassle of sending the camera in for the work to be done.
In fact, in practice, the biggest problem I've had with the 5D3 is the missing battery grip, lack of an RRS's L-plate for said missing grip, and having to run release candidate software (LR4.1 RC) to be able to work on the images.
* Based on my testing (though EV1 is the lower design limit for the 5D3's meter) and I reserve the right to adjust those numbers based on further testing.