I guess I'm not being clear enough. I was trying (and failing, obviously) to distinguish between a naive implementation that did not consider future expansion (the N^3 algorithm I mentioned at the bottom) and an overly-complex solution that did not meet current needs.

To some extent, I guess they are points on an implementation spectrum caused by the same problem...focusing solely on the solution, not the problem.

Thanks for helping me clarify this in my own mind.