I've never heard of a Heisenbug, but that is clearly a play on the Heisenberg Uncertainty Principle [0]. It's unlikely that the author was referring directly to a Heisenbug, instead I'd guess both are different plays on the same name.
It is a play on the Heisenberg Uncertainty Principle as is a Heisenbug, which are bugs that happen differently or don't happen when you are observing the code execute. The typical case is you have a crash you can reproduce, but when you attach the debugger to the process the crash no longer happens due to subtle differences in code execution timing.