If internals are more accurate, then why wouldn't Quinnipiac et al use their methods?
Internal polls are usually more focused on the underlying details - how motivated supporters are, which specific demographic groups are supporting your candidate, etc. The public polls try to do some of that, but don't have the time or the interest in getting too far down in the weeds.
So for example, you could have a "likely voter poll" that shows Trump 25%, Cruz 20%, Rubio 15%. But your internal polling shows that of that 25% for Trump, 10% are low-propensity voters and the Trump campaign has made no contact with them to be sure they get to the polls; while a majority of the Cruz supporters are high propensity voters who have been contacted multiple times by the campaign. That is going to tell you more about who is likely to actually show up and vote than just the top line numbers quoted by the major polls.
Hannity made an interesting point I hadn’t thought of. The GOP candidates ought to be hammering Hillary on what is going on in Iowa on the Dem side instead of this very minor (in comparison to the utter chaos on the Dem side).
It works on a few levels in that it could take her out so they could run vs Bernie, or it re-inforces the corruption of the Clintons and the Dems in general. Lifts all boats.