@Jackdaw, I agree that any FORE-&-AFT angle in the mainsheet will be inefficient, because the boom can't and shouldn't slide fore or aft. But Port/Starboard windward/Leeward) angles are MORE efficient than vertical sheets, since the sheet's job is to pull the boom to windward, to the centerline.
@HEINZIR, you're right, and I see it now! So the Crosby rig is apparently a failed attempt to capture the advantages of a sailing/racing dinghy's split-tailed mainsheet while adding more purchase (mechanical advantage) for bigger mainsails.
So, the question is whether there is, or can be, a SUCCESSFUL attempt! I think I've got one in my head, just offhand: Start with a "Y" spliced mainsheet. (Albacores used to use a standard double-braid Dacron rope with the core pulled out at the "Y", the inside braid going one way and the outside braid going the other way. The "Y" needs to be seized/whipped. Now we mostly splice thin high-tech tails into a thicker base, sometimes all of it non-absorbent $$$ stuff.)
Start with a block and cleat mid-boom and mid-cockpit, with an appropriate purchase. (Much less purchase will be needed than for pure mid-boom sheeting, because downward tension is neither needed nor wanted. The last few degree to the middle will be done by the windward tail.) The sheet goes aft along the boom to a swivel block at the boom end, which also has a pad eye on each side/bottom. Each tail is led straight outboard to a block on the gunwale/sheer, then back to the same-side eye on the boom end, where it ends.
With our simpler dinghy rig, the "Y" splice is positioned so it just barely enters the boom block with the boom centered and high (close-hauled, open leach). With tight vang, the tails run farther forward along the boom. (Boats with thicker tails and smaller (non-swivel) boom-end blocks can experience scary tangles between the loose tail and the tight one.) But I think that WON'T work off the wind with 2:1 tails. But my pre-coffee brain is hurting now, and I need a diagram or a model! For sure, I think it works if the split tails can go far into the boom, though things get messy if they reach the forward (mid-)boom block, or worse, the cockpit block.
And like any keelboat mainsheet that connects at boom end AND mid-boom, this will be a challenge for canvas cockpit enclosures.
Whether or not I've solved this problem with 5 minutes' thought, I can't believe that human ingenuity can't do it, without the complexity of twin mainsheets. But my 5 minutes has made me fonder of the twin-sheet solution than I was at first!