Paleo-baking and sugar exchanges

As Gerry mentioned in his 4th-dinner recap post, I am attempting a paleo-conversion of one of my more elaborate (but so, so worth it) dessert recipes: the (in)famous sour-cherry almond torte. The original recipe has 1/2 cup flour (easy to adapt with nut/tapioca flour) and 1 cup of sugar to eliminate--everything else is mostly paleo-friendly (or at least as paleofriendly as almonds can be).

Sugar, for better or worse, has distinctive properties beyond simple sweetness: it is dry, it browns with baking, and yet more or less retains its character under high temperatures. Most other sweet things do not have this combination of characteristics. Many artificial sweeteners and some natural sweeteners (honey) are not stable under baking conditions. Other natural sweeteners may be heat-stable, but have different physical properties (wetness, for example) that will alter the characteristics of the final baked good. Most often, adaptation of baking recipes to no refined-sugar versions involves replacing the sugar with some exchange volume of apple sauce, molasses, agave, maple syrup, etc., and reducing the volume of liquid ingredients in the recipe to compensate for the added moisture from the sweetener.

Unfortunately, the almond-torte recipe brings with it a significant challenge, as the torte portion has very few wet ingredients (6 egg whites, which provide the leavening for the torte), and yet has a cup of sugar that must be eliminated. I've got some ideas, and I'm open to suggestions (a tip of the hat to Gerry re: the virtues of molasses and tapioca flour). Needless to say, I'll be running at least one test version before Sunday to see what comes out. And if anyone is making mayonnaise this week, I can probably help you out with some egg yolks...