I suspect that the time and cost involved in non-trivial structures will limit any real experimentation with non-traditional designs. It might be trivial to throw away a kilogram of plastic on a failed prototype, but the same doesn't extrapolate up to experiments with bridges.