My calibration results above were with my 30D. I just tried the same calibration with my 40D and the same sheet of paper. I also used the same lens, ISO and aperture, varying only the shutter speed, as before.
This time I started out at +1 and took a series of seven photos, in 1/3 stop intervals, all the way up to +3 stops. Even at +3 stops there was no sign of anything blowing. In DPP I was able to add another 1/6 stop with no signs of clipping and only by adding 1/3 stop to the +3 exposure did I finally see clipping in some of the paper. So I have almost 3 1/3 stops of headroom with my 40D compared to the almost 3 stops of headroom that the 30D gives me in like for like conditions. This is what I said about my 40D in the other thread I referred to earlier.
By the way, the +2 stops exposure on the 40D gave me the same 1/1600 shutter speed that the 30D gave me. So even though the metering is identical, I have an extra 1/3 stop safety margin on the 40D compared to the 40D. I think that is because, according to comments I have read on the net,
HERE and
HERE, 100 ISO on the 30D is actually more like 125 ISO, whereas the 40D is accurately calibrated so that 100 ISO is a true 100 ISO. So the extra "hidden" sensitivity of the 30D can cause it to blow highlights earlier than the 40D. It should also give fractionally brighter images than the 40D when all settings are exactly identical.
What this does mean is that when using my 40D I could meter off the same sheet of white paper but would need to bump my compensation up to +2 1/3 stops for the 40D and only +2 stops for the 30D to achieve the exact same exposure in both cameras. It also shows that you should take the time to calibrate your own camera against your own test target.
Thinking about this further, it doesn't really matter whether you have a 12% grey card, 13% grey card, 18% grey card, a sheet of white paper, or your own palm. The important thing is that you know how much compensation to apply for your chosen exposure target, for your camera.
Finally, if you shoot raw and like to expose to the right, it is important to understand just how much extra headroom you do have above whatever you meter with. For example, let's say you are shooting a sports scene - perhaps kids playing baseball in the park. Consider whether there is anything (important) in the scene that would be brighter or more reflective than the paper. There might be white baseball caps, or shirts, and they would probably be about as bright/reflective as the paper - but what else could/would be brighter still? Maybe some shiny white paint on a car in the background (not very important and should be blurred out of interest by shallow DOF in any case), or reflective metalwork or glass, or maybe nothing. If the answer is "nothing" then instead of only going to, say, +2 stops above the metered reading for the paper you could perhaps go to +2 1/3, +2 2/3 or even +3 stops above the metered reading. You'll still hold the detail in the highlights but will have more raw data to play with and can fine tune shadow detail and/or bury noise if you are shooting at higher ISOs.
If, on the other hand, you are shooting motorsports and could have shiny paintwork, glass and metal as part of your subject, you will likely want to be very careful to preserve detail in those brighter areas, and could happily lose shadow detail under the car or in the wheel arches, for example. In this case you would probably want to play safe with a standard exposure rather than pushing the limits of the right hand edge of the histogram.