Sorry, but I find the idea of someone predicting specifically coda magnitudes to be risible.

Gordon Bennett, when our best predictors are generally having difficulty hitting a barn door at twenty paces, expecting them to discriminate between minor magnitude scales is like asking them to nail a particular knothole. Also, if someone does decide on, say, Mlg (by which I assume you mean mbLg) and even understands what it means, if the agencies report only ML and mb for the next earthquake are you going to calculate mbLg yourself to evaluate the prediction?