This is a rather strange definition of 'learning', a better term would be 'behavioral modification.' Here's a definition of the practice:
> "From an affective neuroscience perspective, our understanding of psychiatric illness may be advanced by neuropsychological test paradigms probing emotional processes. Reversal learning is one such process, whereby subjects must first acquire stimulus/reward and stimulus/punishment associations through trial and error and then reverse them. (2009 Cambridge Pyschological Medicine, Dickstein et al.)"
This is pretty low-level stuff, e.g. 'learning' that red means stop and then 'relearning' that red means go, etc. The authors claim to have developed evidence that LSD is associated with plasticity in healthy humans suffering no psychological disorders.
It's an interesting study, I suppose. Psychedelics are often called 'mind-expanding' and it's possible that the underlying neurochemical mechanism of action facilitates breaking out from fixed behavior patterns. This might also explain some of the reasons why authoritarian governments don't like psychedelics as they might encourage a loss of faith in authorities, or cause religious fundamentalists to reconsider their belief systems etc.
After all, if you can't necessary trust your sensory organs to give you reliable information about the world around you due to the influence of a complex molecule on a receptor protein in your brain, perhaps blind faith in the pronouncements of politicians, priests, and media talking heads is not justified either?
In terms of learning, however, I've always thought that most notable psychedelic effect was the development of good three-dimensional visualization of complex objects, such as electrical fields or proteins and DNA and things like that. Such visual effects are generally associated with higher doses however, which are a bit tricky to manage safely, and so require special attention to optimal set and setting.