Could friction be reduced by instead of using (1) 1/2" spacer, using (4) 1/8" spacers? When putting 2 delrin (black) spacers next to each other, do you do round end to flat end or round end to round end, do you put the flat side or round side against the bearing block? Also, do you use teflon washers everywhere that there are shafts? How about using teflon washers instead of spacers?
From a physics standpoint, replacing 1 spacer with 4 should make no difference, the normal force ends up being exactly the same. Intuitively, that would be less efficient. Round to flat vs. round to round doesn’t make a difference either. You very likely won’t see a difference with any of those things you talked about unless you compress the things on the drive shaft somehow. I would advise against going with the 4 spacers, though, since the parts aren’t all the same. Tolerances might work against you.
Complexity is exactly correct. It will have no effect.
I would say that complexity is right, in that you will notice little to no difference, but I would suggest using the biggest spacers that fit so that you have spacers that fit in smaller places when you need them. If you used 4 smaller spacers in place of 1 big spacer, you won’t be able to use said big spacer for smaller spaces, so it is more cost efficient to just use the big spacer I believe.
That is the same rule of thumb I use when building. Use the biggest spacers were possible so that you save the little ones for when you need them.
My favorite one is the white Teflon washer I use them all over the place even in places I could use a metal washer.
I would like to know friction tests on a few surfaces… specifically:
Teflon washer vs Teflon spacer vs black spacer on metal and bearings.
Also shaft collar on bearings.
These would be useful tests but probably would not make much of a difference on how your actual friction.
Well if the coefficient of friction between two spacers is less than the spacer-metal interface, then theoretically there would only be slip between the two spacers, reducing friction. I don’t know if this is actually true - in fact, the spacer-metal coefficient could actually be lower - but it might work. Given that, it would still not be worth it to have more than 2 spacers in any given application, as you don’t reduce the effective coefficient of friction any further and just add complexity.
Teflon definitely has the lowest friction, second is delrin (black spacers), third is nylon (white spacers). Teflon on metal is just as good as teflon on delrin. However, if it is lubricated, the material doesn’t matter much, but you don’t need to lubricate teflon. Lubrication in the bearings, where there will likely be the most pressure and therefore friction is even more important.
http://www.roymech.co.uk/Useful_Tables/Tribology/co_of_frict.htm
this has some very interesting information about friction, particularly the generalizations stated in the first section.
What if the coefficient of friction at the teflon-teflon interface increases as the surfaces heat up? Then having as many teflon-teflon interfaces as possible would keep any one of them from heating up too much and reduce your overall cof 
You will want to use the largest spacer if possible, as previously mentioned. If heat is an issue, then you will have to deal with it as you go. In the meantime, use the largest spacers. It is more efficient.
I don’t think that using fewer spacers will result in a performance increase, it’s just better practice because it means you’re less likely to run out of parts. But yes, heat is not an actual concern. My comment was supposed to be tongue in cheek :o.
I think by “more efficient” he meant “more efficient use of parts.”