First, I apologize. This is a reposted version because the previous one had factual errors.
This question follows up on a previously raised question.
Let ${\bf A} \in {\Bbb R}^{n \times n}$ be a symmetric positive definite M-matrix whose non-diagonal entries are non-positive. Let ${\bf B} \in {\Bbb R}^{n \times n}$ and ${\bf C} \in {\Bbb R}^{n \times n}$ be two diagonal matrices, whose diagonal entries are non-negative.
$${\bf M} := \begin{bmatrix} \mathbf{A} & \mathbf{0} \newline \mathbf{0} & \mathbf{A} \newline \end{bmatrix} + \begin{bmatrix} -\mathbf{B} & \mathbf{C} \newline \mathbf{C} & \mathbf{B} \newline \end{bmatrix}$$
Can it be proven that the ${\bf smallest}$ eigenvalue of ${\bf M}$ always decreases with the diagonal entries of ${\bf C}$ increase? (Which has been verified numerically)