And here is a photo in total:

https://fstatic1.mtb-news.de/v3/26/2654/2654118-c5wk34oi19ex-img_4645-large.jpg

Perhaps an experiences builder will see the twist here already? Or am i to picky with the twist?

Thanks!

Marcus

   Have a beer and be proud of your work!  I looked at the photos, even expanded on my computer screen, and can hardly see the twist, and only in a couple of them.  I had to go back to your post to make sure you were talking about mm and not cm. Until you get into the 5mm plus or even the cm range of measurements I think you're fine, and probably have less "twist" than the majority of these boats ever built.  I built a Chess 17 and remember checking for twist, but all I did was lay two sticks across the hull, one towards the bow and one towards the stern, stand back a few steps off of each end of the boat, and make sure they looked just about parallel.  So I don't really know if I had (or have) a few mm of twist in the boat, I never tried to get that accurate.  And the boat works wonderfully.

If you're really worried and can figure out a way to make it happen, you might be able to frame up, strap, whatever the boat to take out the twist while you're putting in the deck beam and nailing on the deck.  That might be a method of taking out the twist. Might be tough to figure out how to do this, and might run as much risk of causing more distortion rather than less, but is at least one idea for fixing this, even though I don't think there is even the slightest reason to think you don't already have a near-perfect boat.  When you put on the arched deck beam and deck you'll probably be using a lot of strap pressure and pulling in and out on the sides of the hull anyway, so I don't think a few mm of "twist" that you're looking at right now really means much of anything at all.

 None of those are much. I  suspect the boat will be a 3 to 4 mph boat average speed that really won't be affected by a 5 or less mm difference in the gunwales as much as your paddle techinque. 

I would rig a center line bow to stern and drop a plumb bob at both ends to see what "twist" shows up in the bow and stern of the boat. I suspect no much. 

That method assumes that if the gunwales are off the rest of the boat is off. But it might be just the angle of one side panel of the hull.  To be completely over concerned take measurements off of the reference center line to equal stations on the chine and see if they are equal. Then do the gunwales at the same station.  Then disregard. 

I can't interpret your laser level stuff. I'm used to checking for twist the way one does when hand planing planks. The 2 sticks are like winding sticks. The picture that comes closest to telling me anything is this one:

If that was my boat, I'd move the sticks wider apart and then just stand in front of it at the centerline, back up at least a boat length, close one eye and bring my head down to the point where the tops of the sticks line up. If they're parallel, there's no twist.

The beauty of doing it that way is that it doesn't matter if the boat is level or not. This method only cares about the relative positions of the sticks and your eye, and it doesn't need additional hardware that could introduce more errors.

The reason to move the sticks further apart is to increase the sensitivity of the test. Unless you've managed to build the boat into a corkscrew shape the twist will not reverse and disappear over the length of the boat.

And as the other guys say, it it's only a handful of mm at the ends of the sticks, no one will ever notice and it will make no difference. This is a kayak, not a hypersonic missile.

Have fun,

Laszlo