I agree with you. Many of the NP-complete problems are appealing because they don't seem difficult to solve at first glance. The fact that there are simple phys

I agree with you. Many of the NP-complete problems are appealing because they don't seem difficult to solve at first glance. The fact that there are simple physical processes that do a good job for small cases doesn't surprise me, one can solve many of these problems by hand for small values of n. But, O(e^n) makes the big cases intractable even if the small cases are very very easy. It is the behavior of the algorithms as n gets large that is important and demonstrations of physical analog methods to solve modest cases doesn't seem relevant for the reasons that you mention.