Dyno1 is not an accurate depiction of what the numbers really are they are inflated. This can be done by over inflating tires, where the tire sits on the drum, placing a fan where it shoots directly into the a/c, pulling the bike at lower temperatures after the oil is hot, playing games with how the clutch is handled. A myriad of tricks can be played to inflate or detract these numbers.
It also seems as though as Dynojet continues to make changes to their software/algorithm that the numbers seem to be dropping a little each update, which is why some shops won't upgrade to the latest or near latest software/versions.
one dyno could have the latest software and another be older is one reason, model of dyno's can differ as well as manufacturers and type. Such as Dynojet, Mustang, Superflow, etc.
This is why some vendors use X dynoshop, its not just a dyno tune and a result, it is used as an advertising tool, unfortunately you get sold on X set of numbers only to find out they are inflated. this isn't a deal killer if your getting a good tune then its just a realization and recommend a verifying pull somewhere else to level set the results and expectations. Higher numbers helps sell parts / bragging rights and entice people. When used as a tool to define how the bike runs and that the a/f is right or in back to back application it can show gains and losses and be comparable.
In this case you have 1 stealership with a happy dyno and a 2nd shop with an honest dyno based on the numbers provided. The results of the dyno if no games were played shouldn't be a sole determining factor in how their work is, although it would make one question what version there on, when it was last calibrated, how good the stack is performing on the dyno etc.