Just a small comment on the rotation discussion.

Breaking masts was common in the Tornado class when the spi was introduced, before the sailors learnt to handle it. Keeping the downhaul on, to depower, was a really bad idea. Letting the mainsail out was also a quick way to go mast-shopping.
After a while, sailors became confident and began experimenting with mast rotation and sheeting in various windstrengths. Now, we rotate the mast to get a nice profile and sheet the mainsail actively. That is, until we start trapeezing, then we rotate the mast fully and the skipper begin using the traveller instead of the mainsheet. Top teams would probably still use the mainsheet instead of the traveller, but we can not afford a new mast so we keep it safe. This is with the relatively slender Marstrøm mast.
It's faster with less rotation and an active mainsheet, but you need to weight it up against possibly breaking your mast, as Wouter says.

PK, when talking about a deep spi, how deep is your spi and how flat do you want to go? The current Tornado spi's are about 20% in the deepest part, and they build a lot of virtual wind. With a flatter spi, you can go higher on the run and build more virtual wind, but you will probably hurt in low-wind conditions when you can't build enough virtual wind to justify the extra distance. Rick White sails with a "hooter" (I would call it a reacher), a spi flat enough to use roller furling, and swears by it. He also makes it go well in weak winds, but others have mixed experiences with it. It's up to you, but it's questions and experimentation that makes the boats go faster. Then again, out of 10 ideas you try out, only 1 will make you go faster, and that is also expensive


I am currently look at spi shapes, and are pondering the 'vertical' component of the spi's resultant force (force from the sail is generated perpendiculary off the surface, so you get a small vertical component). As we are a heavy team, we dont need much help keeping the bow up, except in the heaviest weather. If I could re-shape the spi so we could get a larger forward driving component, we could go double trapeezing downwind in a blow instead. Need to weight this against possibly backwinding the mainsail badly in the top, so it's all about compromises as usual. And of course, double trapeezing with a heavy team downwind might break the mast, which brings us back to the current topic..