The total path length TPL sums up all vertex depths. As such, it is related to the Sackin index, i.e., the total external path length, and the total internal path length. TPL is defined for arbitrary trees and it is an imbalance index, i.e. for a fixed n\in\mathbb{N}_{\geq 1} it increases with decreasing balance of the tree. It can be calculated using the function totPathLen from our R package treebalance.

Definition (e.g. Dobrow, Fill, 1999, Takacs, 1992, Takacs, 1994): The total path length TPL(T) of a tree T\in\mathcal{T}^\ast_n is defined as TPL(T) :=  \sum\limits_{v \in V(T)} \delta_T(v) = S(T)+TIP(T),
where S(T) denotes the Sackin index, i.e., the total external path length, and TIP(T) denotes the total internal path length.
Computation time: For every tree T\in\mathcal{T}^\ast_n, the total path length TPL(T) can be computed in time O(n) (see Fischer et al., 2023, Proposition 23.5).
Recursiveness: The total path length is a recursive tree shape statistic. More information on the recursion can be found in Fischer et al., 2023, Proposition 23.6.
Locality: For arbitrary trees T\in \mathcal{T}^\ast_n the total path length is not local. For binary trees T\in \mathcal{BT}^\ast_n it is local. (see Fischer et al., 2023, Proposition 23.7).
Maximal value and (number of) trees with maximal value on \mathcal{T}^\ast_n and \mathcal{BT}^\ast_n: An upper bound for the total path length for arbitrary and binary trees trees T\in \mathcal{T}^\ast_n or \in \mathcal{BT}^\ast_n with n leaves which is tight for all n\in\mathbb{N}_{\geq 1} as well as the (number of) corresponding maximal (binary) trees has already been established (see Fischer et al., 2023, Theorem 23.3).
Minimal value and (number of) minimal trees on \mathcal{T}^\ast_n: A lower bound for the total path length for trees T\in \mathcal{T}^\ast_n which is tight for all n\in\mathbb{N}_{\geq 1} as well as the (number of) corresponding minimal trees have already been established (see Fischer et al., 2023, Proposition 23.8).
Minimal value and (number of) minimal trees on \mathcal{BT}^\ast_n: A lower bound for the total path length for binary trees T\in \mathcal{BT}^\ast_n which is tight for all n\in\mathbb{N}_{\geq 1} as well as the (number of) corresponding minimal trees have already been established (see Fischer et al., 2023, Theorem 23.4).
Expected value and variance under the Yule model: Open problem.
Expected value and variance under the uniform model: Open problem.