The logical extreme of this is that golfed code contains the fewest bugs.

You missed the ultimate silly extreme. No code means no bugs. Try and disprove that :)

Taking good ideas to silly extremes does not disprove them as good ideas.

For a huge proportion of the population -- of almost anywhere -- eating less is a good idea. That starving yourself causes death does not disprove that.

I think that "less code means less bugs" is a nice catchphrase, but that noone has actually proven this.

Actually, they have. Over and over.

When FORTRAN and COBOL took over from assembler, it was clearly demonstrated that more code was written more quickly with considerably less bugs.

And given equivalent skills and testing regimes, coding in Perl will result in less bugs than writing in C.

And if the anecdotal evidence isn't self-evident enough for you, I'll try and dig up references to some of the research I was party to back in the '80s. Though I'm not sure I'll be able to access it.

In the mean time, try looking in the Haskell world. It's one of their mantras and they usually back their stuff up with research.

You missed the ultimate silly extreme. No code means no bugs. Try and disprove that

That's trivial to disprove. Most code is intended to do something. If it doesn't do what is intended to be done, it isn't correct. No code means nothing is done.

trivial to disprove.... If it doesn't do ...

Sorry, but your disproof has bugs.

What "it". If there is no code, there is no "it".

The absence of functionality is not a bug. In many cases, it is the most pragmatic, correct solution.

---

Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error.

"Science is about questioning the status quo. Questioning authority".

In the absence of evidence, opinion is indistinguishable from prejudice.

Please show me the bugs in my code above.