I agree 100%. Plus, the pause and repeat method actually requires keeping the language in memory for more than a few microseconds. In fact, we can often get better results by deliberately extending the time between listening and repeating, sustaining the audio memory in our short-term memory as long as possible. I honestly don’t know why so-called “shadowing” is such a buzzword. By the way, it's rather easy to go through the motions of shadowing without actually achieving accurate pronunciation (which means the student is practicing bad pronunciation). From what I've observed, the best first step is to study the phonology of one's target language before trying any sort of trendy technique.