The same exercise requires us to prove that $\lim_{x \to \infty}x\log(1+\frac1x)=1$ just before proving the identity in the title. I can use this fact as follows:
$$\lim_{x \to \infty}\left(1+\frac1x\right)^x=\lim_{x \to \infty}e^{x\log(1+\frac1x)}=e$$
Is this proof valid? I'm asking because the author is using a complicated $\varepsilon-\delta$ proof which could have been avoided, if I'm not mistaken.