Moreover, his influence on states all over the world has been largely counterproductive: constitutions jammed fill with checks and balances and heavily separated powers are no less unlikely to be led by tyrants than states without; the idea that all branches of government jealously guard their own privileges against the other leading to a neat equilibrium is one that is not seen in reality.
I'd actually be interested in seeing this put to the test. Of course, the central pitfall of such a research design would be to take constitutions at face-value; we would not only have to control for socioeconomic realities, but also the actual "correlation of forces" as of the constitution's ratification.