You are expecting something that doesn't exist.
I'm confused. I have a lot of expectations but as far as these templates go, none whatsoever about points -- only about stock rails. Even there, I'm much less concerned with the internal geometry than in making sure that all three tracks, as they exit the turnout, are oriented properly with respect to one another. These are layout planning templates, not tracklaying templates.
I realize that a turnout might be drawn simply as a triangle; I've seen it done. I thought I could do a little better than that without trying to reproduce every detail. My main concern is the track centerline; that *must* coincide from one section to another. Another issue is locating curved portions accurately so that clearance can be maintained as well as possible. Car body overhang will cause train-to-train interference any time tracks are too close together and trucks are not aligned -- I don't see how these rules can be suspended through turnouts. (I'm speaking of clearance between a turnout and an adjacent track. Fouling between diverging and straight routes is another issue.)
Difficulty: Drawing the centerline accurately enough to ensure the eyeball doesn't lead me astray. It's not enough to lay one section on top of another and fiddle around by eye and say, "Yeh, that looks tangent." I rotate and displace sections arithmetically.
Only a portion of the diverging route is curved; the question is *what* portion. I've made most of the diverging route straight (everything from the end to about an inch pointsward of the frog). If I make another portion straight, that will do nothing to justify the Fast Tracks data; the remaining portion will have to have an even smaller radius.
Meanwhile, sorry, but you seem to be contradicting yourself. I'm sure I just fail to understand. But if the turnout flexes the points (they're not hinged), then surely they'll be curved? It is clear that there is no such thing as "a" #8 turnout. Every prototype, every model manufacturer, is different.
I'm not a fan of slavish imitation of prototype. The model should look realistic but I will generally trade faithful reproduction for reliable operation. I'm definitely building HO, not Proto87.
As I said, I can't criticize Fast Tracks turnout in operation; I lack experience. It's the documentation that's driving me up the wall.
Does anybody *use* Fast Tracks? If so, could you make a physical measurement of gauge at several locations throughout the built turnout and see if there's some explanation of the stated 0.66"?
I guess the criticism I would find most helpful at this time is of the practicality of my current set of templates. If I design a layout around them, will it be buildable?