The hyperbolic tangent and generalized Mellin inversion (Q1806248)

The objective of the author is to establish an interesting identity concerning the hyperbolic tangent function \(\text{th }x\), and then apply it to simplify the formula for the Plancherel measure for the symmetric space \(\text{SL}(n,\mathbb{R})/\text{SO}(n;\mathbb{R})\). The identity, for the case \(n=3\), becomes the well-known relation: if \(\alpha+\beta+\gamma=0\), then \(\text{th }\alpha \text{th }\beta \text{th }\gamma=\text{th } \alpha+ \text{th }\beta+\text{th }\gamma\). Main lemma: Let \(a_{ij}\) \((1\leq i\leq j\leq n)\) be such that \[ a_{ij}=-a_{ji}\quad \text{and}\quad a_{ij}+a_{jk}+a_{ki}=a_{ij}a_{jk}a_{ki}.\tag{1} \] Then \(\prod_{1\leq i<j\leq n}a_{ij}= \frac{1}{2^\nu\nu!}\sum_{\sigma\in S_n} \text{sgn}(\sigma) \prod^\nu_{\ell=1}a_{\sigma(2\ell-1)\sigma(2\ell)}\), wherein \(S_n\) is the symmetric group, and \(\nu=[n/2]\) (the integral part of \(n/2\)). The proof is by induction on \(n\) and straightforward but somewhat involved (pp. 695-705). The relations (1) permit the simplification. Instead of the product of \(\frac{n(n-1)}{2}\) factors \(\text{th} \frac{\pi(t_i-t_j)}{2}\) that appear in the Plancherel measure we have a \textit{sum} of products of \(\nu=[\frac{n}{2}]\) factors, and it turns out that the resulting integrals are all equal. Consequently \[ d\mu(t)=\text{const.}\prod_{1\leq i<j\leq n}(t_i-t_j)\cdot\prod^\nu_{\ell=1}\text{th} \pi(t_{2\ell-1}-t_{2\ell}). \] Remark: In [Dokl. Akad. Nauk SSSR 133, 503-506 (1960; Zbl 0097.10903)] and his dissertation (Moscow University, 1961) the reviewer computed the Harish-Chandra \(c(\lambda)\)-function for \(G=\text{SL}(n;\mathbb{R})\), \(\text{Sp}(2n;\mathbb{R})\) and the orthogonal group \(O^+_f(2n,\mathbb{R})\), wherein \(f\) is the quadratic form \(f=x^2_1+\cdots+x^2_n-\cdots-x^2_{2n}\). These involve also the products \(\prod_{1\leq i<j\leq n}\text{th} \frac{\pi(t_i+t_j)}{2}\). Question: Is a simplification possible also for \(\text{Sp}(2n;\mathbb{R})\) and \(O^+_f(2n;\mathbb{R})\)? This question has relevance to the problem of making explicit the Selberg trace formula for \(\text{Sp}(4;\mathbb{R})/\text{Sp}(4;\mathbb{Z})\).