Not entirely. Ample paging space may enable you to fit more processes into RAM - even if no actual pageouts occur - while maintaining system stability. This has to do with the nature of demand paging.

I'm not sure I follow you. How does swap space help demand-paging page in things not swap-related? I know Windows requires every bit of mapped memory to be backed by swap, but didn't think UNIX had this limitation...

No idea about how Windows work but Unix, at least both modern Unixes & Linux do not require RAM to be backed by swap, hopefully.

HP-UX has indeed this bizarre approach by default and is the only current Unix lacking virtual swap. This design limitation can be partially overcome by enabling pseudo-swap.

Overcomitting memory is a different and dubious beast.

You are correct however Scrutinizer and I somewhat derailed the discussion toward HP-UX and Solaris behaviors.
Linux, which is definitely on topic in this thread, is different as it overcommits memory. That means it doesn't care if allocation can be fulfilled when initially asked for by the program, unlike the Unix versions we were talking about.

Trivial to disable. echo 2 > /proc/sys/vm/overcommit_memory ; echo 100 > /proc/sys/vm/overcommit_ratio

But it's useful, as just demonstrated; there's plenty of environments that don't have disk space to throw away like that. And somehow the world never ended from it despite all predictions No OS can run anything without sufficient memory, in the end it's just another strategy.

You are right, overcommiting can be disabled with recent enough Linux kernels.

Note that if you do it, you won't be able to efficiently use all of the available RAM in your live-cd Linux configuration.