Geometric explanation for the “−1” offset in the pluriclosed root-parameter relation
Determine whether there exists a geometric explanation for the appearance of the “−1” offset in the relation x_{α+β} = x_α + x_β − 1 that characterizes left- and Ad(T)-invariant pluriclosed Hermitian metrics on a compact semisimple Lie group G with respect to a fixed maximal torus T and root system Δ^+. Specifically, explain why these pluriclosed metrics satisfy x_{α+β} = x_α + x_β − 1 for all positive roots α, β, α+β ∈ Δ^+, whereas the Hermitian structure (J_{𝔮}, g_{𝔮}) on the flag manifold G/T is Kähler precisely when x_{α+β} = x_α + x_β without the “−1” term.
References
Intriguingly enough, the last line in the theorem should be compared with the K\"ahler condition for the Hermitian structure $(J_\qg,g_\qg)$ on the flag manifold $G/T$, given by $x_{\alpha+\beta}=x_\alpha+x_\beta$ for all $\alpha,\beta,\alpha+\beta\in\Delta+$, whose solutions are precisely $$ x_\alpha:=\sum\limits_{i=1}{2d} k_ix_i, \qquad\forall\alpha=\sum\limits_{i=1}{2d}k_i\alpha_i\in\Delta+. $$ We do not know whether there may be some hidden geometric reason for this behavior.