Agree with discussion of issues with tolerances. Not only does the horizontal spacing between injectors have to be exact, but lateral position, angle and height also all have to match. I would expect there is some degree of variation between injectors and between engines/head casting, especially over a length production run.
The expansion difference is probably not enough to cause an issue. The difference in expansion coeeficents between al alloy & cast iron is about 6 micro inch/in *f... this means a total difference in length change between the rail and head of about 0.00015" per *f of heating.... so for a 100* temperature increase you'd expect to see the rail expand about 0.015" more than the head. Whether this is enough to cause issues is hard to say. However, the return rail is probably going to be somewhat cooler than the head since it's got a lot of area to lose heat to air flow and relatively little heat flux in from the fuel so the difference may actually be less. One could also engineer in this expansion when designing the rail and spec the injector spacing to have the holes centered at operating temperature. That being said, the o-rings are likely flexible enough to handle the variation. It is possible that the repeated compression/release cycles could effect the long term life of the o-rings though. In any case, I expect that variation from manufacturing tolerances would have more impact than the differential expansion.
A more compelling concern to me though is that this doesn't do anything for the primary issue. Generally, the cause of the leaks is old stiff O-rings. This will continue to be a problem regardless of if it's a plastic fitting or metal. If anything I think the rigidity of the metal rail and expansion issues would put more stress on the o-rings and make them more likely to leak. I would also question the surface finish of the o-ring seat in the rail. If this surface is rough, which is may be depending on how it is machined, then I would expect this to also potentially cause shortened o-ring life.
IMHO, the only issue with the plastic return line fittings is damage from external issue - ie you break them working on something else. I could be wrong, but I don't think there's much issues with them spontaneously breaking. Likewise the rubber return lines have a reasonable life expectancy. Yes, it's not infinite, but they last pretty well. Easy and cheap enough to fix as well. Beyond that, you still have the rubber hose to fuel filter, injection pump, and back to the tank to potentially leak.
IMHO, not worth the cost.
/tangent
I also hate the use of the word billet.... that's the most overused marketing term ever. And highly misused.
