Tom, thanks also for your remarks. I want to address your "I don't think there will ever be a plain language version of QFT", because I think there is a possibility, and indeed that section III of my essay here /may/ be a starting point for at least a slightly more accessible account. The problem is not the basic formalism, insofar as I claim that QFT is at heart just a form of stochastic signal analysis ---notwithstanding the incompatibility of joint measurements at time-like separation---, the problem is renormalization, which makes the literature and textbook accounts almost entirely obscure. The difficulties of understanding QFT are notoriously extreme, but a lot of work has been done to make Feynman's dictum that they are insurmountable less true than it used to be.
It's in the nature of the evolution of our understanding of mathematical formalisms that they are gradually reformulated, and that each reformulation leads to a wider understanding, until a popularizer at some level takes hold of the new ideas and brings them to wider audiences. Whether it's my work or someone else's that brings QFT into clearer focus is not of course of the essence.
Looking at your FQXi essay this season, I see that you have detected at least something of this process in the Bell literature. If you haven't previously seen my "Bell inequalities for random fields - cond-mat/0403692 (v4, 24th May, 2006), J. Phys. A: Math. Gen. 39 (2006) 7441-7455", perhaps you might find my approach there to /those/ questions a little different from much of the Bell literature (though people also find that paper opaque). IMO, it's just a question of time, perhaps another 5-10 years (or 10-20, ...), before the penny drops for the Physics community that for a wide variety of reasons the violation of Bell inequalities proves almost nothing. At some point it will become possible to say that of course most Physicists always understood that was the case, which it will be possible to justify by pointing out that 't Hooft, Bohm, and other Serious Physicists were never outright refuted, and the derision nay-sayers were subjected to will be forgotten except by Historians. After the paper above I decided to stop worrying about Bell inequalities and move on to thinking much more single-mindedly about QFT.
If you want to understand QFT, I'd recommend trying to engage with Haag's "Local Quantum Physics" for a few years, though it's definitely not easy going. Amongst the standard textbooks, I find the wildly old-fashioned (yet still modern enough) approach and notation of Itzykson & Zuber far more congenial than Weinberg. The path integration of modern textbooks, in particular, obscures the algebraic and combinatorial structure that is the only path to reconciling QFT with ordinary QM, IMO. That doesn't remove the ultimate need for at least some people to understand the relationships between such different approaches, of course.
Although I operate at not much above the level of popular accounts, I believe that a lot of what can be shown about complex systems is probabilistic in character, right? I suspect the relationship between probability and statistics will be forever difficult.
It's becoming more worthwhile having posted here. Thank you.