I believe the main reason for the 1Plus2 not being offered as a custom-fit in the end was because of the fact that it would have to be tuned for tubing. Acoustic tubing has one major challenge over a chambered system: quarter-wave resonances due to standing waves that reflect in a semi-closed system that need to be tamed (or in some cases, amplified). Tubing is normally chambered at a constant diameter, which means that sound waves are neither compressed nor distorted. However, it's often a goal of acoustic engineers to use bell/horn-like designs or Helmholtz resonators to modify the frequency response of a transducer system. The 1Plus2 was designed with that open, reverse horn system in mind. It would take a lot more effort to match the same type of response with acoustic tubing; it's not impossible, just takes a lot longer, and probably a lot more thinking (about tube diameters and lengths).
Regarding the differences between demos and custom-fits, I see the number one difference between comparisons is improper wear (leading to lessened isolation and/or insertion depth), and poor tip selection. Demos are just more difficult to obtain a proper fit, because they're not built for your ear precisely. The effect of silicone tips plays a factor in resonance damping and amplification in the upper frequencies, but can be minimized with the proper choice of tips, by using tips that don't protrude beyond the opening and that don't create a reverse horn effect (i.e. needs to be wide enough).
Does bone conduction factor in? Perhaps. I used to think it was a lot more important, but now that I have a far better understanding of the way acoustics work in the ear, I have to agree with James in that bone conduction plays a much larger factor in the microphonic effect. Large area sound waves hit the skin of the head, and concentrate toward areas such as the mastoid process, sending sound waves up toward the bony part of the ear canal. Back reflections emanating from the exit ports of a custom-fit or universal-fit IEM contribute relatively little. If anything, bone conduction plays a process in the occlusion effect more (though the occlusion effect technically also has to do with the salpingopalatine fold and the resonance of the eustachian tube).