Implement an MC simulation to compute passband integration errors

[kbwestfall]

The passband integration errors, used for both the emission-line moments and the spectral indices, are likely very poor. We've known this for a while given MC experiments from Portsmouth and analysis of the repeat observations provided by the Beijing team. For #27, I did an idealized test of this and I don't think it's an error in the formula used for the error propagation, but the test shows again that the errors are bad.

We should implement an MC in the DAP for these error computations.