1. The conditions in Assumption 4 is unnecessarily strict. We only need $\mathbb{E}[\abs{\mu(u_\ell(t)) \, f(t)}]$ to be bounded instead of the square one. This is clarified in my thesis.