it's inefficient for a turbolift (no matter how small you make it) to make at least two curves during its descent/ascent.
Why would the lift have to be efficient? It's unlikely the pylons have been designed to convey turbolifts; rather, they have been designed to perform some other job, after which the turbolifts will have to cope. The same with the neck of a TOS Constitution: a straight one would be better than an angled one, and a turbolift will have to do at least one S turn to clear the neck from top to bottom, or then travel diagonally.
FWIW, the "Hero Worship" Oberth had an internal layout consistent with a size of about 120-150 meters, unless the decks were higher than usual.
Also, the pylons didn't seem to have anything on their "decks" save for the floors. If the turbolift just travels down "feet first" but tilts enough to follow the curving of the pylons, it will be going down a shaft that can be up to two meters wide for a 120 m ship - wider than the standard turbolifts of TOS and TNG - and will only have to negotiate one shallow curve at the bottom and then flip 90 degrees. If it travels down "side first", negotiating the curve will be slightly easier, but another 90 degree flip will have to be introduced at the top end.
Timo Saloniemi