When I put the rails on my NED there were some significant bending forces, and I used rope & etc. on the extended fore and aft "tails" of the rails to hold them in place along with clamps - lots of clamps, maybe one every 8 inches - but once all glued up everything was fine, the tails were sawed off and nothing like this remaining force separating the plywood laminations was noticed.  Weird.  Maybe you got some extra stiff rails? 

A really quick suggestion off the top of my head is to consider steaming the rails in place, maybe getting them to take a set in their currently bent condition, relieving the stress.  You could do this with wet towels and a heat gun.

And if your rails are well-glued to the transom and breast hook I'm suprised that there is enough remaining force that can be transmitted to the plywood strake to delaminate it.

And of course at some point getting some epoxy down into the plywood laminations where they are separated and keeping them clamped while that hardens.  Some ring-nails or screws in the delaminated areas running from inside the top strake out into the rails might add even more strength to the repair.

Bubblehead has good repair advice.

Did you build from a kit or from plans? Something seems quite unusual. Once the second rail is glued to the first, and the epoxy is given time to cure fully, the curve should be maintained (mainly) by the glued joint in the rails.

At the bow you can leave some screws permanently in place into the breasthook. You might wish to countersink them a little and then fill with wood-flour epoxy. You might be able to get a screw in at the transom too.  

Thanks to you both.

I'm building from a kit.  The first rail had 24 hours in clamps then many days to cure.  The second rail had 24 hours in clamps.  I've got one rail on the other side, with no indication of the plywood failing.

I understand that these are essentially glued laminations, and that they should hold their curve once glued together.  It seems to me that there is still some residual tension (spring-back) in each piece, and that the tension in both pieces added together is enough to start peeling the plywood apart.  (If this were a piece of furniture I wouldn't expect a bent lamination with two layers, each ~7/8" wide and 7/8" thick to hold in stiff wood....I'd use more like eight 1/4" thick layers.).

It also looks to me like the rails supplied in my kit are sapele, which is substantially stiffer than the mahogany or cedar that the manual says is usually provided.  Maybe that has something to do with this unusual behavior.   I'm guessing here, and could be way wrong....I'd be glad to be corrected.  Especially if there's an operator error option I'm missing.

But three out of four rail pieces are on, so I believe it's a matter for forging ahead, not backing up.  3" screws into the breasthood is certainly do-able, although screws parallel to plywood laminations doesn't strike me as strong joinery.  And I can put some screws through from the inside in the stern.....ugly, but maybe unavoidable.  That means whatever tension there is will be transferred to the butt joint between the panels and the transom.....I hope the fillet can take the stress.

,,,c