I personally don't think interactions between humans or processes that can be performed mentally or verbally should be patentable. I don't think you should be capable of infringing a patent by sitting, thinking, and talking to your friend.
The MoT test excludes these things and is the best test we have to date on excluding these types of processes. Is it the perfect test? Probably not. But it's damn near close. The problem is the courts can't just tell examiners to use their own discrection on interpretation of 101, they have to always create some "test".
I think it should be possible to infringe a patent by sitting, thinking, and talking to your friend but only provided that whatever portion of the activity that infringes is new, non-obvious, enabled and definite in view of the prior art methods of sitting, thinking and talking, and provided some new function or advantage is provided.
If I come up with a way to increase the productivity of human interactions mentioned above, the question is, does Public Policy want for me to keep my mouth shut about it, or encourage me to disclose it by providing me a limited monopoly so that everyone can ultimately benefit ? How about, if I know how to reproduce what Ed Leedskalnin built at Coral Castle, using my mental powers alone, and I know of a way to teach it to you ? I might disclose it if I could patent it, but if I can't I'll take the secret to the grave with me and you all can build your own danged pyramids using those inferior mechanical contrivances.
It should be an interesting case, that Bilski opinion. Tests aren't always clearcut, but they're better than nothing. I think that uniform application of all such "tests" is important, and have yet to see all examiners apply Graham as it was put forth some 50 odd years ago.