Scapular positioning assessment: Is side-to-side comparison clinically acceptable?

Scapula positioning has been considered a very important component of a competent shoulder examination.  Assessing scapula positioning can tell us about possible decrease in the subacromial space depending if the scapula is anteriorly tilted at rest, for example.  However, when dealing with overhead athletes, is scapula positioning ever "normal"?  This study, from the Manual Therapy Journal, aimed to look at the reliability of assessing side to side comparisons of scapula positioning in healthy, young adults.  
What the study produced was that the scapula on the dominant shoulder showed greater retraction(P<0.001; (p)=.68) and upward rotation(P<0.001; (p)=.70) at the three resting positions in the study (rest, hands on hips, 90 degrees of shoulder abduction with IR).  However, the interesting part was during shoulder abduction from rest to 90 degrees, retraction, upward rotation, and posterior tilting were similar bilaterally.  This raises the question: With our overhead athletes who typically have different resting positions of their scapulaes, how much do we try to change?   
Now, we also have to take into account this study sample was relatively small (14) and that previous studies have suggested that scapula positioning is extremely important (see Reinold, 2009 for example), but it is something that we should ponder.  When assessing our overhead athletes do we know if there was scapula dyskinesias before injury or was that their normal positioning due to their dominant arm?  

Any thoughts? Any experiences?  Feel free to comment below.  

---