The way Craig frames his position in his opening remarks about Scrooge is important. The spirit of Christmas future shows Scrooge what would (emphasis Craig) happen, given certain conditions. The critical point is at the 14 minute mark: Craig says God knows what "people would freely choose" in any circumstance "in which they might be placed" logically prior to decreeing in what circumstances people are placed.
This is the essence of Molinism. It's almost like God runs a simulation of what "I" would "freely" choose in scenario x, y, and z. God then determines to create the conditions for scenario y so that the simulation plays out in reality. In this way, the Molinist maintains that God is in control of all things. After all, God could have run scenario z if He wanted to, so He is "in control." Yet the Molinist will also claim that people are free in a libertarian sense, since "my" choice in scenario y was (allegedly) free. God's knowledge of what my choice would be in scenario y is not dependent on His decree that scenario y actually happens - just the opposite.
The problem with this is referred to as the grounding objection, and as I understand it, it can be explained in various ways:
If God's decree is "logically posterior" to His knowledge of what I would freely choose in a given scenario, the question is begged as to what "grounds" God's knowledge of what I would freely choose. God allegedly has middle knowledge, but how? It can't be grounded in God's natural (necessary) knowledge, because we aren't talking about necessary truths or realities. It can't be grounded in God's decretal knowledge, since God's decree is (allegedly) logically posterior to God's so-called "middle" knowledge of what I "would" "freely" choose in a given scenario.
So on what basis does God know what I would choose in a given scenario? A simulation? A simulation is never the reality. The question is how God can know, if I am free in a libertarian sense, that I would choose a definitively certain way given scenario y. For until the scenario is "run" (played out in reality such that I am able to make a choice in that scenario), I haven't actually made a choice (I am not the same as a simulated version of myself).
A Molinist will likely demure and say that the basis for God's knowledge is not a simulation. But since middle knowledge cannot be something tied to God's own nature (i.e. things dealing with necessary truths or reality) nor God's decree (i.e. things dealing with God's free choice), then it seems all that is left is that God's "middle" knowledge depends on a quasi-real version of myself. Perhaps a more neutral way to frame the point is by asking how I can be really free (in a libertarian sense) if God knows that I would act a certain way logically before I actually do.
Even if the Molinist continues to protest against these representations, the main point (that even the Molinist would seem to admit) is that God's "middle" knowledge depends on something external to Himself. Even Craig admits this when he elsewhere says, God "has to play with the hand He has been dealt" (link). This statement that Craig made in the past gives away his and the Molinist's position: it seems that someone else has dealt the cards and, therefore 1) God is not in control of all things (e.g. the dealt hand), and 2) God's knowledge is dependent upon whoever or whatever dealt these cards.
This is severely problematic. If God's knowledge is eternal (as I think and have argued it must be) yet dependent, then what is the nature of that on which His knowledge is dependent? Is that something else co-eternal with God? And aside from this question, is it possible for God to be eternally omniscient if His knowledge is dependent upon an extrinsic source (I would and have argued in the above link is not)?
This is the main counterargument against Molinism. Now, as for what Craig says about Molinists and Calvinists being like-minded in agreeing God has overriding purposes for moral evil yet there being a significant distinction in that for the Molinist, God can "will" in given circumstances that one choose good whereas a Calvinist cannot, his argument is equivocal. Calvinists can maintain the same distinction Craig would between God's moral or prescriptively will vs. His decretal will. In a given circumstance, a Calvinist can maintain God commands other than He had decreed.
Furthermore, for the Molinist, God does not merely permit moral evil, He decrees it. The Molinist would agree God could have not created at all. So the decree for God to create this world means that the Molinist must agree God wants moral evil to occur. Whatever reasons He might have for decreeing a world in which creatures freely choose evil, He wanted this world.
So much, I think, is the standard Calvinist reply to Molinism. I think the following illustration, however, highlights the point. In contexts involving the problem of evil and the Reformed response, I tend to hear a lot about God as the "author of evil." But consider two publishers:
1) Publisher 1 publishes a book he wrote, and the book he wrote includes characters that choose moral evil.
2) Publisher 2 has drafts of books handed to him (by whom? who knows!), and he chooses to publish one in which characters choose moral evil over any drafts of books in which no one chooses moral evil.
Is there a significant moral difference between the two publishers? I fail to see one. One might argue that because publisher 1 was also the author of the book, he is morally "responsible" for causing the characters who choose moral evil in a way that publisher 2 was not.
Now, a publisher who is also the author has more direct, intentional impact on why characters act a certain way. But his reasons for writing them the way he does need not differ from the reasons a mere publisher perusing the same book might choose it involving the same characters. If it is morally acceptable for the second publisher to publish the same book as the first publisher (who also authored the book), the burden is on the one who argues that the first publisher acted in a morally unacceptable way to pinpoint why the first publisher's more direct and intentional connection is morally unacceptable. Simply stating that the first publisher was also the author is not an argument against the moral acceptability of his actions.
In fact and on the contrary, I would argue that an author who intentionally writes characters a certain way is, if anything, in a better position to know why the book is worth publishing in contrast to a publisher who has been handed several drafts ("dealt" a hand) and left to divine - pun intended - which one tells the best story!
So while Molinists like Craig may charge Reformed Christians with believing God is the author of sin - and I have argued elsewhere that whether or not this is accurate or even problematic depends on what the person making the charge means by "author" - Molinists like Craig believe God is the publisher of sin. For the same reasons Craig would not view his position as problematic, Reformed Christians have analogous appeals as to why their view is not problematic. A book that is worth publishing was a book worth authoring.
