Thinking more about the McCalman selection, I believe it all hinges back to Deans wanting to start with Smith. He is throwing the dice a bit here, and good on him for doing it (to an extent) but there is a lot of risk. If injuries and fitness hold up, then we will win, if they don't (ala Brisbane) then we will lose.
For the Brisbane test I said my biggest worries were 1) Injuries and fitness and 2) JOC (James O'Connor) at 10. It proved to be our downfall (although the Lilo one was very unlucky).
For this test Deans is making the following Risk/Reward gamble:
Reward - Starting Smith gives us more grunt at the ruck, more experience, better 'on ball' skills and great linking ability. Furthermore if Smith can last 80 then we can double team them with Hooper, moving Smith to 6 or 8 (a big starting pack and a fast finishing one).
Risk - There is a good chance that Smith (unlike Hooper) can't last 80 minutes, so if we want a replacement for Palu then we have to carry an extra forward on the bench for Smith (ie McCalman). I suspect McCalman will only get on if Smith really starts to struggle - Deans will prefer to have Hooper/Smith/Mowen finish the game. The risk is obviously the cover for the backs and it is increased due to the knowledge that AAC (Adam Ashley-Cooper) is carrying an injury. If one single back gets injured we have to rely on the uncapped and somewhat inconsistent Mogg to get us home (either at 15 or on the wing).
This whole risk/reward equation could have been negated by starting Hooper and brining Smith on for Palu, however I think Deans lineup will work this best provided there are no injuries to the backs.