Erica, I think I can help clarify some of your questions. I have a lot of experience operating both powder and deposition printers. While I haven't used a ProMetal setup, I can take an educated guess that maybe could be further 'tempered' by Glenn if he stops in.
I would say, do not sprue metal parts. You are just going to have to accept that very small metal items will raise their price somewhat. I have items that were only a few cm3 before and they did not change much at all. I still think it is a great deal when we are talking about custom, short run production. Sprued metal parts are very likely to break because of the rapid change in thickness. That part will cool first, and the difference in expansion between other parts will cause it to crack. The labor to distinguish, then grind and polish those sprues cannot be given for free. The stem has outlived its use by this stage and might break, but it doesn't matter. When you take into account your labor of removing your sprues and the fact that you won't be able to offer other finishes, it pretty much rules that out for Stainless Steel. Because of the extreme likelihood of failure, I would reject all sprued metal parts if I were reviewing models for Shapeways, except in very specific cases, and they would need a very experienced person to recognize those cases. The best economical action is to order many at once, which you already do.
From a depowdering perspective I can say that removing a part with exposed fine details from a powder bed can easily take ten times more labor. I would also reject parts like that, regardless of the fact that they barely hit the minimum wall thickness, otherwise the cost for all parts would have to rise. I always think about these issues when coming up with the design. My Time Keeper model has 1.8mm elements, but it also has a thick ring around the outside that makes it easy to pick up between the thumb and forefinger and brush it off without breaking it. Also, all of these thin elements are supported after distances of no more than about 1.5cm, and are always supported on both ends, making them far less likely to break, and they are thicker in the direction parallel to how impacting forces are most likely to be applied. In retrospect I regret not explaining all these things to people because some very unlikely designs were created that simply focused on the numbers in the design rules, when the reality is obviously much more complicated.
All in all, it is the attempts to stretch the design rules and get around pricing structures that led to the dramatic price rise for tiny items. This became pretty obvious several months ago after the FUD tests. I know people were not trying to abuse the system, it's just that our attempts to make ultra-economical models broke the Shapeways pricing system. I have to take partial responsibility for doing things that raised peoples expectations without explaining exactly how and why a particular design works. I'm doing my best to rectify that but there is nothing I can do to hide the fact that the reality is indeed complicated and takes some thinking and work if you are going to squeeze out every penny. Please don't describe it as Shapeways "Forcing" you to cheat the system, because as we have seen, that will simply break the pricing model and force the prices even higher, when what we should be doing is recognizing which behaviors are causing the problems and keep it in mind when we create designs moving forward.
3D printing does allow us to do amazing things and the price is far better than any other alternative, but we are going to have to move just little farther from the limits if this is going to work.