Furthermore, on those occasions when you *can* transfer information from one part of a system to another, thermodynamics limits how much information you've got without putting more energy into the system. (and I mean real work-equivalent physics energy, not made-up woowoo mystical energy.)
Example: put two pieces of magnetic tape next to one another. Imagine one is a recording of, oh, say, Bing Crosby, and the other is blank. Leave them there for a loooooong time. So, if you listen to the second piece of tape, and sure enough, you're going to hear a faint recording of, guess what, Bing Crosby. But guess what else: the original recording has degraded by at least as much. The information transfer is two way, and while one tape is impressing Bing Crosby on the other, the other is *erasing* the first! (This actually happens, BTW, it's a hazard of magnetic tape stored on reels for a long time.)
To make a copy of a tape without destroying the original, you have to use an amplifier, and an amplifier requires energy. That's not just an engineering problem, it's fundamental to the nature of information.
So *even if there was* a mechanism to exchange information about therapeutic substances from one sugar pill to another (and there isn't any coherent explanation that I've ever heard of), the originally inactive pills would be impressing the state-of-not-being-active onto those pills that are activated (whatever the hell that means in homeopathy). The originals would degrade. If you're talking information theory (as some homeopathy "theorists" attempt to do), this isn't additional dilution, its a degradation in the quality of the "signal" and would therefore be expected to eventually destroy the effectiveness of those pills.
Now, one could argue that this is why homeopaths are encouraged to go to the supplier to get fresh remedies, but this does not explain why homeopaths seem to feel its ok to use grafting, or why its ok for the supplier to use grafting themselves. Rolfe's point is very relevant and bears repeating: either grafting works, and you can always do it, or it degrades, and you cannot rely on the effectiveness of a grafted remedy (or indeed the original remedy you grafted it from!) and therefore shouldn't graft.