The problem with testing our economic equilibrium accurately is that it requires people to actually care about the currency. That means the stake has to command a certain level of real world legitimacy and confidence for the voting behavior to be indicative of anything.
This is why I'm against testing it out with SMTs and other currencies that have yet to be really established.
Smart traders paper trade their systems first, and then find that reality behaves differently. This is why large quant funds do back-testing first and then forward-testing with small allocations before scaling to meaningful size.
I go back and forth on this a lot. I think test nets are important, and SMTs will give an opportunity to test a lot of things. On the other hand, the most important impacts will only occur if the change happens universally across Steem instead of being opt-in.
Long-term, I don't think it all matters that much. SP holding will be about RC's instead of vote value, and most of the value distribution will be SMT's within communities. We do have to actually survive to the long-term.
While I responded negatively to your other comment about the tax, I think you are correct about the overall impact (and need) for these economic changes.
I'm also skeptical that we can survive through to the long term without a reasonably functional economic system in place for our base token. As a pure bandwidth token, many other projects have beaten us to it, and can do what we do faster and better.
I think test nets are better used for technical tests of function than economic behavior. We're interested in the latter which can only happen if people really really cared and are playing with real money. Maybe I'm being insufficiently imaginative with SMTs that can be pegged to certain other currencies etc, but I don't see people caring about them for a long time for voting behavior surrounding them to be a useful indication of anything.
I'm all for SMTs, and I think even if these economic changes are implemented alongside them they'll take no more than a weeks worth of extra dev time.