Evidence from cognitive neuroscience supports action observation as part of an integrated approach to stroke rehabilitation

40 pages trying to explain how observing an action helps stroke rehabilitation. I wonder whether this is still applicable since mental imagery was not found to be helpful;
http://oc1dean.blogspot.com/2011/04/mental-practice-with-motor-imagery-does.html
And where do mirror neurons fit in this possibility?
http://oc1dean.blogspot.com/2011/05/impressionable-brain-and-mirror-neurons.html
the 40 page article here:
http://download.journals.elsevierhealth.com/pdfs/journals/1356-689X/PIIS1356689X10001141.pdf
the first two paragraphs with interesting parts highlighted;
This review of cognitive neuroscience research aims to show
that observing meaningful actions, for example, using a rowing
machine (Fig. 1), can contribute positively to the stroke rehabilitation
process.
Following a cerebrovascular accident, individuals may be left
with chronic motor impairment (e.g. hemiplegia), and cognitive
and psychological disability. Fortunately, recent advances in brain
imaging have led to a greater understanding of the mechanisms of
post-stroke recovery. This knowledge has been essential for optimizing
the efficacy of interventions aimed at promoting motor
recovery. All forms of neural reorganisation are possible following
stroke and include: synaptogenesis; diaschisis, sprouting from
surviving neurons; and recruitment of functionally homologous
pathways (Rossini, Calautti et al., 2003) Therefore, any functional
post-stroke intervention should attempt to enrich and optimise
neural stimulation in order to promote this brain plasticity.
Unfortunately, many traditional therapies have tended to focus on
the recovery of prescriptive physical functions and on the
resumption of generic activities of daily living and, in so doing, have
neglected previously valued activities which have personal
importance to individuals (Cott, Wiles et al., 2007)In my case canoeing. Research that
has attempted to address some of these concerns, by considering
the effectiveness of motor imagery after stroke, has suggested some
promising results (see Page, 2010 for a review). However, a number
of practical and procedural concerns still remain (see Holmes
2007). Until recently, video presented practical problems for the
amateur user; cost, size of equipment, technological understanding
of hardware and editing software were all barriers to its use. The
developments in ubiquitous digital media now allow for the quick
and easy production of high definition images that can address
imagery generation and ability problems, and provide relevant
contextual information. As well as the methodological benefits of
this approach, action observation has also received support from
the neuroscience literature.