The most accurate way would be to build a physical prototype. Obviously, in many cases this is the whole reason Flow Simulation was purchased, so that wasn't necessary.
If you are looking objectively at the problem, you can bracket the results. You know the air will be no cooler than 31.9C. You can create an upper bound case by assuming no heat loss to the air and rerunning the analysis for "heat conduction in solids" and this should give you an upper bound on the solid temperature.
You can use the information on Convective Heat Transfer to do a simplified hand calculation to see if you are in the same ballpark. After all that, you should know how comfortable you are with your results.
First things first - have you done a simple control volume energy balance? Check the average temperature of the air at your outlet and see that it carries away the 115W generated inside the volume. After that you can worry about if the pattern of flow around your batteries makes sense.