Here are some overexposure tests for white levels of the BMPCC, as well as some documentation of technical issues with the BMPCC that might complicate the issue (or might not, this is not my expertise as I am doing this at the directive of a developer).
Here are the the raw files for white point calibration, shot with a tungsten light source for smooth spectrum lightsource, filming while moving the light closer and further away to change exposure. I shined it in a way to make a slight gradient in light level across the white paper I was shooting for this test. The subject was out of focus, so that any fine detail could be attributable to camera sensor and image pipeline artifacts.
Although the request is for one 1-2 stop overexposed raw file, I decided to give you multiple, because I noticed a few issues that might complicate figuring out white points. First, I noticed a weird fixed pattern grain in highlights very close to or at clipping (not 100% shure as I don’t have Rawdigger, the most accurate way of assessing), and it is observable even when it is safe to say the files are being correctly decoded (camera original files pixel peeped in Davinci Resolve). Note this is not an issue that warrants trying to do lots of trouble shooting to fix, as it is only visible at 100% or more with excessive sharpening, and is likely pretty easy to fix in most pro NLE video editors. I also noticed that the rgb channel levels, according to the waveforms in Davinci Resolve (Again, disclaimer, this is not Raw Digger), seem to reach a saturation point (Example file labeled 4 Saturation light.dng), with an observable variance of the levels, depending on the pixel. As the overexposure rises further, the flat lines of levels rise just a little bit, as well as getting just a little less variance in values (5 Saturation Heavy.dng).
https://drive.google.com/open?id=11gbxOLWAj7mMo5AXGX9VpcgFfGADXUM3
I also posted this on the Github issue.