Of little application here perhaps, but I have built a series of models with winglet tips (on an F1A glider). They were made to the Whitcome pattern. I tried the De Grantz tiplets but the added complication of the fore tiplet was not worth the effort.

Findings -

Properly made they work. The twist from root to tip is very critical.

The washin/out setup is likewise critical. I used -9* at root which gave -1* at the tip. (Yea that is how much twist they have!!)

You need to give careful consideration to the flow from wing undersurface to tiplet. As this is the transition of the tip vortex, the positioning of the tiplet and its function in this area is also critical. You can not just glue them on top of the wing.

They are extremely vulnerable particularly in collision with fences.

How do I know they work?

Knock them off one side and then try and tow the model. It will turn towards the side with no winglets. Quite strongly.

The effect on the model? The glide was a slow and shallow (estimated at 6 to 10 inches) fugoid. No sharp stall. Best estimates of cause was that the winglets were on the very edge of effectiveness.

There is an Italian who has been experimenting extensivly with tip vortex control on F1C models. A search on the 'Net should turn up his work which is thorough and has produced unusual results.