The whole premise of your calculation is wrong. The specific % of hands you minraise is not a direct function of your opponent's strategy because each hand in that % has different EVs. Your approach has led to blunders like including the expectation of open folding in your computation for the expectation of minraising--clearly this is wrong since once you've minraised, you can't open-fold.
Instead you should be looking to find cut-off hands--that is find hands that are very similiar in EV minraising and open-folding, and then all hands worse are clear folds and all hands better are clear opens. Total up all the hands you're opening and you will find your open %.
With your assumptions of no flatting and 50% 3-bet shove you'll find that all hands that minraise/fold have the same EV minraising as open-folding (though in practice there will be card-removal effects that will change this a bit...). All hands that can call a shove will be strictly better minraised then open folded. This means your strategy can open 100%, or open just your min-raise call hands, or something in betwen--ev of these strategies will be the same.
That being said villain not flatting is a horrible assumption