I was there when the no swap policy was under lots of active discussion. It was there to ensure that application authors could have strong expectations on how long a memory access would take. Imagine trying to provide evidence that removing the "no swap" policy wouldn't affect any of the applications that were running across a very large infrastructure.
Her point really resounds though, this person was referred to all the time in conversation, "well, we probably couldn't get this past X". He was mythical. In general that was often a good thing. Making changes that could have such a broad impact should be well reasoned. But it's also quite annoying, having a nagging voice inside your head that just says "no", seemingly disconnected from the actual source.