Every aspect of telescope design is a compromise that can be optimized. For example, consider aperture. More aperture means deeper (and narrower angle) viewing but at the cost of size, weight and price. Is a larger aperture scope used occasionally in dark skies a better value than a smaller scope set up in the yard? Perhaps the answer is both but then budget takes a hit.
Two unheard of optimizations that I make are a narrower upper end and intrusion into the light path of the coma corrector. What do I gain by doing so?
A narrower upper end means less weight, a smaller spider, perhaps a smaller diagonal and potentially a correspondingly smaller mirror box. In a fast focal ratio telescope, the diagonal is typically sized to deliver adequate (not 100%, but 70% or so) illumination at the edge of the lowest power eyepiece. Looking from the extreme off-axis point, we see that the portion of the primary mirror that is vignetted by a narrower upper end is very small, in the neighborhood of several percent or a few hundredths of a magnitude. It is worth it to give up a few hundredths of magnitude illumination at the extreme edge in order to go down in diagonal size.
Optimizing diagonal size is all about selecting for the smallest responsible diagonal. Moving the focuser closer in to the diagonal means a smaller diagonal. For fast focal ratio scopes, the minimum diagonal size is calculated by dividing the diagonal to image plane distance by the focal ratio. For example, if the diagonal to image plane distance is 12 inches and the focal ratio is f4, then the minimum sized diagonal is 12/4 = 3 inches. For very fast scopes, f3.5, f3.0, f2.5, the coma corrector can jut into the light path. So the focuser is moved outward, resulting in a larger diagonal. But let's think about what we are truly optimizing, namely, obstruction. Obstruction reduces the light going to the image plane and produces diffraction. More obstruction, more diffraction (the shape of the obstruction determines the shape of the diffraction). The total obstruction of a minimally sized diagonal and the coma corrector slightly jutting into the light path can be less than a larger diagonal by itself. This is a worthy optimization to make, minimizing light loss and diffraction.
A narrower upper end and allowing some intrustion of a coma corrector into the light path are two optimizations to consider.