So, Mark Witton had a lengthy article on why paleontologists think the Quetzalcoatlus northropi could fly.
One interesting thing about it is that Witton and Habib think the Quetzalcoatlus had a 90-second window of anaerobic wing flapping to gain altitude, then transitioned to soaring, with glide ratios similar to storks.
Now dragons could do something like this, for everything else there's magic. However, they have to carry an extra pair of legs and a tail, and most likely a larger torso (though you can have most of it filled with air).
I don't have any concrete numbers, so we'll go for the maximum:
- The muscle fibers need the highest power output, with the least mass, i.e: a high specific strength
- How much of this muscle fiber can be packed into the dragon, assume similar flight mechanics (muscle and tendon placement, etc) to birds
- Tendons have the same tensile strength as spider silk, approx 1.3 GPa, bones are reinforced with goethite fibers, I tried to find a compressive strength, but with no results
- The flight muscles will only be used for 90 seconds
So, yeah, I basically ask if it's possible to create muscle tissue through biological processes (and a quite long time span) that can fulfill these criteria.