Actually, there is no need to make guesses about whether cdaddr is bad, unreadable, introduces hard to handle bugs or what the limits of list structure complexity should be. It isn't some new idea or brief historic oddity that you can pass whatever judgement on.
The term is at least 40 years old, standardized, still in use, has its own established pronunciation and isn't known to cause a lot of bugs. It is also at the periphery of the car/cdr body of terms -- i.e. programming tradition shows that the level of complexity it covers is needed (or the term would have disappeared), but any further is too rare to merit a shortcut. In addition, it is easy to remember what it does: each d is a cdr, each a is a car. So it isn't any harder to read than a chain of firsts, seconds, rests and nth n's.
The term is at least 40 years old, standardized, still in use, has its own established pronunciation and isn't known to cause a lot of bugs. It is also at the periphery of the car/cdr body of terms -- i.e. programming tradition shows that the level of complexity it covers is needed (or the term would have disappeared), but any further is too rare to merit a shortcut. In addition, it is easy to remember what it does: each d is a cdr, each a is a car. So it isn't any harder to read than a chain of firsts, seconds, rests and nth n's.