The questions are from Naive Lie Theory by John Stillwell, problem 1.5.2.-1.5.3.:
My main question is about 1.5.3., but I believe I should explain how I understood 1.5.2. For example, if we want to rotate about $P$ by $\theta$ in the diagram below, we first reflect by line $M$, then line $L$ will be flipped to the other side of line $M$. Then we reflect through this flipped line $L'$, then we get the rotation about $P$.
But I could not understand how 1.5.3. works. If I rotate about $P$ and then about $Q$, the point $R$ is moved from the original point (either by using reflections or just simply rotating them). It is because once $Q$ is moved by rotation about $P$ so that the length from the moved $Q'$ and the original $R$ becomes longer than the original distance between $Q$ and $R$, no matter how we rotate about $R$, $Q$ cannot be superimposed at $Q'$. Therefore, I could not see how the composition of the two reflections become a rotation about $R$.
I believe there is some mistake in my understanding but I could not figure it out myself. It will be helful if I could actually see each step of the proposed composition of rotations.

Write the two given rotations in terms of reflections at lines present in the image. Then observe that some cancellation takes place.