*via Stephanie O' Neil, whose brainchild it was to make this list, and who has several entries overlapping with the above
† via Paramita Dutta
‡ via Raghav SG with the help of Claude AI
^ via Ashutosh Jogalekar
Steve Martin adds: Maxwell used "D" not because it stood for electric Displacement, but because he chose to denote quantities introduced in his paper by letters in alphabetical order of appearance. (Of all the quantities he defined, D is one of the few that actually could have stood for a sensible word, displacement, because he wrongly believed that it represented the physical displacement of an aether. But he never said "D stands for displacement".). Since he just started at the beginning of the alphabet for all the quantities he wanted to define in his paper, we ended up with "A" for what we now call vector potential, "B" for what we now call magnetic field, "C" for what we now call current (this choice was later rejected by Heaviside, who standardized many of Maxwell's other choices), "D" for what we now call electric displacement. He used "E" for total energy which again accidentally makes sense, and not for what we now call electric field. Instead, he used "P,Q,R" for the x, y, z components of what we now call the electric field E. He also used "a, b, c" for the x, y, z components of B. The whole thing was a notational disaster.
Flip Tanedo adds: The [paucity] of 'd' and 'o' symbols for words that do not start with 'd' or 'o' reflects that physicists aren't completely insane: they know that 'd' is to commonly a differential and that 'o' looks too much like zero, and so we should *not* use them as symbols. (Unless, of course, there's a compelling reason to use that letter, e.g. 'd' for distance or 'D' for diffusion coefficient' or occasionally 'O' for observer.)
I'd never heard of the coefficient of restitution, but 'e' for anything other than the base of the natural log seems like a faux pas. Similarly, I am begrudgingly okay with capital I for current since it is different from lowercase i for the imaginary number. (I feel similarly about script-I/scri for null infinity.)