This proof below really seems to have little to do with uniform integrability, what does the DCT application give exactly?
Any alternative proofs would be good too

This proof below really seems to have little to do with uniform integrability, what does the DCT application give exactly?
Any alternative proofs would be good too

Copyright © 2021 JogjaFile Inc.
The calculation in the proof shows that
$$\begin{align*} X_t := |M_t N_t - \langle M,N \rangle_t| &\leq \sup_{s \geq 0} |M_s| \cdot \sup_{s \geq 0} |N_s| + \sqrt{\langle M \rangle_{\infty}} \sqrt{\langle N \rangle_{\infty}} =: X \in L^1. \end{align*}$$
Consequently,
$$\int_{X_t \geq R} X_t \, d\mathbb{P} \leq \int_{X \geq R} X \, d\mathbb{P}.$$
Since this implies
$$\sup_{t \geq 0} \int_{X_t \geq R} X_t \, d\mathbb{P} \leq \int_{X \geq R} X \, d\mathbb{P},$$
the claim follows by letting $R \to \infty$ (using dominated convergence at the right-hand side).