Patrick,
The first thing you need to do is carefully inspect your crank for any other problems (to establish what other expense may be incurred). There are a couple of inherent faults that restrict the life of Dragonfly cranks. The first is the centre web - it is drilled it's full length, so has a weak spot across the hole which often results in a hairline crack that releases the grip on the crankpin. This may be the reason you have excess endfloat on the conrod - ideally, it needs to be about 6-8 thou. If the crankpin has moved it should be evident when looking at the end of the crankpin from the other side of the centre web. If the crankpin hasn't moved, the excess endfloat may be due to wear on the faces of the webs/conrods. The second inherent fault is that Douglas tried to improve the bigend lubrication by feeding oil direct through drillings in the crankpin (instead of sending the oil to the centre web and splitting it 2 ways and feeding the side of the bigend). With the 'improved' design, if the endfloat of the front conrod is not tied up accurately, the centrifugal force of the spinning crank pushes all the oil out of the front bigend - thus starving the rear. As a guide to whether your crank has moved - given a tolerance of plus or minus 5 thou - the centre web, and the gaps between the webs should all be .650".
When I was reconditioning cranks, I bored the rods and fitted needle roller bearings with hardened outer tracks - these outer tracks have smaller I.D. than the original rods, so made it possible to establish the correct endfloat even if using worn rods or crank webs. As an experiment, when I rebuilt the crank for my Dragonfly, I reverted to the older system of oiling, using undrilled crankpins and a Mark type centre web that went back to feeding the oil to the sides of the bigends. After that crank had done 30,000 miles, I checked it out and found it to be as good as the day it was first fitted - as they say, 'Proof of the pudding'!.
Hope some of this helps,
Regards,
Eddie.