The problem with greyscale images from the publisher's point of view is that the file sizes are much greater, which is reflected in the number of CDs in the sets, making them more expensive to publish. The poor quality of the pages also means that many of them have to be rescanned at different settings to capture the maximum detail. The greyscale images of the 147 Yorkshire Pieces HO 107 /1210 - 1357 fill a massive 44 CDs, whilst the B&W images covering the same Piece numbers fit on just eight discs. However those larger file sizes should in theory manage to capture every little bit of available detail from the films, optimising their readability.
My overall impression of the new S&N CDs was that the greyscale images were much more readable. This didn't mean that it was all legible though, some pages still had unreadable areas. This is unavoidable given the condition of the originals and the National Archive films. I thought I should be a little more scientific about my comparison, but there was no way I could compare every enumerator's book in all the 147 Pieces for Yorkshire!. My solution was to take an enumerators book at random from every Piece with a number ending in zero, in the hopes of achieving a reasonably representative sample. This sampling method seemed to work as the resultant books contained examples of both the best and poorest examples I had come across.
I then viewed every page in both sets of books and noted whether it was fully readable, sometimes with difficulty, or had any portions that couldn't be made out with any certainty. In some cases this was the entire page, but in the main it was only a portion of a page image. Having gone through both the B&W and Greyscale images I found that the S&N greyscale images offered over 30% more fully readable pages than the B&W images. However this percentage increase in full readable pages doesn't fully reflect the fact that the Greyscale page images, mostly suffered from small areas of poor clarity, whilst on many of the B&W images, the areas were generally much larger.
Unless the National Archives could re-film the census using better technology, we have to accept that the 1841 census is always going to give researchers problems, whether using microfilm or digitised CD images. However I am totally convinced that the greyscale images produce far more readable data than B&W images of the same material. Given the choice between B&W or Greyscale images of the 1841 census, I would certainly choose the latter as in my opinion the greyscale images offer the best of a bad census.