Question on Input file exposures process

I have a Canon 5D (Original version) that can only provide 3 AEB exposures of 2EV steps each. I expect to take pictures with higher DR, thus will make manual exposure "brackets" via shutter speed increases (dark to lightest) in series.
I have tried to translate the few guideline documents into English, and have read them through carefully, but still have a couple of questions.
1. is it recommended for the darkest exposure, for that exposure to be set such that it is exposed to the right (ETTR) such that no highlights are blown out, and just within the right side of the histogram, and then use 2EV brighter shutter settings to progressively lift the exposures until the darkest areas are well within the left side of the histogram?
I think I understand the calculations for the number of images needed for 2EV steps.
So my question is really...is there any benefit to taking any darker images than what I described above, or have I understood the recommended process correctly?
2. in terms of image quality, if the process means I need to use, say, 3-5 input images to cover the full DR of the image, is it best to use RAW input images, or TIFF, for the very best image quality?
3. Is it better or recommended to sharpen the individual images and do any noise reductions PRIOR to the SNR-HDR process, or best to leave the input images exactly as they are, and adjust afterwards? What about white balance/color changes, etc?
I have tried to translate the few guideline documents into English, and have read them through carefully, but still have a couple of questions.
1. is it recommended for the darkest exposure, for that exposure to be set such that it is exposed to the right (ETTR) such that no highlights are blown out, and just within the right side of the histogram, and then use 2EV brighter shutter settings to progressively lift the exposures until the darkest areas are well within the left side of the histogram?
I think I understand the calculations for the number of images needed for 2EV steps.
So my question is really...is there any benefit to taking any darker images than what I described above, or have I understood the recommended process correctly?
2. in terms of image quality, if the process means I need to use, say, 3-5 input images to cover the full DR of the image, is it best to use RAW input images, or TIFF, for the very best image quality?
3. Is it better or recommended to sharpen the individual images and do any noise reductions PRIOR to the SNR-HDR process, or best to leave the input images exactly as they are, and adjust afterwards? What about white balance/color changes, etc?