The counts() function can take an ExonCountSet object and return the raw or normalized counts. The easiest thing to do, then, would be to c <- counts(ecs, normalized=T) and then just apply(c, 1,...) appropriately on the results.

Edit: I should have browsed SEQanswers first. Apparently this is even easier to do with estimatelog2FoldChanges (presumably `getOnlyEffects=T). I'd never noticed that option before!