The boat used for the experiment was a Kittiwake ketch from Swallow Boats.  This has a standing lug-sail and mizzen, and sails well in the lightest of winds.  The foremast is about 10 feet long and the gaff (the top-most spar) is almost as long, hoisting the sail peak very high and presenting a respectable area of sail to the wind. Click on the image to see it enlarged.
The prototype crabclaw sail was made of "polytarp", stretched as tight as possible between two flat wooden spars of equal length.  The spars are strung to a flying mast made from a sawn-off broom handle, and the whole thing is hoisted on a shortened mainmast of about 6 feet . The sail is much lower down, and the prototype sail is probably about 2/3rds of the area ideal for a boat of this size.  The mast may also be too far forward.  However, with careful adjustment, it sailed to windward quite well.
But sailing to windward requires you to "tack" across the wind.  The prototype does this in a novel way - over the top of the mast.  The manoeuvre is very much like flying a two-string kite.  In fact, if you let go of the two sheets, the sail will naturally settle in a fairly harmless position over the top of the mast like this.  This rest position provides some shelter from the rain and the sun, an additional benefit of the rig.
You complete this particular tacking manoeuvre by pulling down on the port sheet and freeing up the starboard sheet. In practice, the two sheets in the prototype are a single line you effectively pass through your fingers.  And then it looks like this. Because the sail is too small for the boat and (therefore) positioned too far forward, the prototype needed additional lines to control the precise position of the apex of the sail.  We think some of these should be unnecessary if the sail is correctly balanced and sized, but we'll only find out when we build a bigger one. Watch this space.
The main part of the experiment was to see whether we could "tack over the top" of a short, unstayed mast.  That bit worked.  But we also discovered some other interesting things about a sail which can be "flown" in several positions.  This is the sail going down-wind in light airs.  With the wind behind, most of the power of a sail is aerodynamic drag, which in this setting depends mostly on the area presented to the wind.  But there is also a wind gradient, and with the wind behind you, in light airs, it's good to have as much of the sail as high as possible.
Sailing downwind with a conventional sail - even a crabclaw - runs the risk of a so-called "jibe" where the sail moves violently from one side of the boat to the other.  With a crabclaw sail flown symmetrically as shown here, there is no chance of this happening.  This is a forward view of the sail going downwind, and as you can see, you can vary the angle of attack to control the forward power of the sail very precisely.  Of course, at the shallow angle shown here, the sail could start to generate some genuine lift - i.e. start to lift the bow of the boat. This, again, may be no bad thing with down-wind sailing where there is often a natural tendency to "bury" the bow.
Finally, by rejigging the forward control lines, we found it was possible to fly the sail downwind with its apex aloft.  This might be a better configuration for very strong winds.  In principle, this setting would generate even more pronounced vertical lift, since the crab claw would be operating in its most efficient lift-generation attitude with respect to the following wind.