Also I have pointed out in my video and the follow up posts in this thread about a meta issue that destroys the Nash equilibrirum, that for the case where there are external failures (external to the block chain's perspective of itself) due to external actuation of cross-partition state, the Nash equilibrium fails because the entire coin fails
If all inputs to a script arrive via one transaction, you must be forced to conclude that there is no difference between said scripting blockchain and an asset transfer chain in terms of equilibrium breakdown?edit: actually, I see what you're saying - if the script author's cause a cross partition dependency, that dependency can affect the resulting ordering on partition merger. The problem is when this dependency is not visible to the system.
further edit: however, I still struggle to see why this would result in anything more than the failure of the script in question?