If you refer to the 'sand-box' algorithm titled "Bounding-box PCA fit", then it's a very simple algorithm based on Principal Component Analysis. And it doesn't work very well (this is why it stays in the 'sand-box' section). On some clouds it can give interesting results, but on others it will be very bad (because it simply looks for the dimensions along which the cloud is the most 'elongated'. This works particularly badly on clouds with a rectangle shape (as the most elongated dimension is the diagonal, and not the width as you would expect).
I don't see any difference with the algorithm described in the post... Have you tried both algorithms on the same dataset? If yes, how much do the two results differ?
Daniel, CloudCompare admin