Justin,
I am assuming a custom bike with a geometry the way you want it. If you keep the seat tube angle and head angle constant, the length of the seat tube will not affect the rider's position.
On a Rivendell, the top tube slopes, and the head tube is extended. Certainly, the sloping top tube will be shorter (measured along the axis of the tube) than a horizontal one will be (as the sloping one takes a "beeline" to the seat tube). Also, if the head angle is steeper than the seat angle, a top tube higher up will be longer. But these effect are small, and in my experience, negligible.
On bikes that interest me, I check seat angle and reach: for the latter, I measure the distance (c-c) seatpost-handlebars, parallel to the ground. This takes into account the stem as well. Note that with higher bars, you need more reach.
Finally, the reason for Rivendell spec'ing their bikes with sloping this and extended that is that most buyers are so used to getting bikes that are too small (by Rivendell standards), that they'd balk if they were told the "true" size of their new bikes. A modern rider who is used to a 54 cm "compact" frame would be hard-pressed to accept a 60 cm frame for their new bike. So Rivendell suggests a 56 or 57 cm frame, and everybody is happy.
I've been there. And today I am glad that my 58 cm Rivendell in fact really is a 61 cm frame, if you assume a (virtual) horizontal top tube that starts at the top of the headlug extension.
That said, I prefer the looks of my 61 cm Singer. Both ride the same.
For an illustration, check Vintage Bicycle Quarterly No. 3, p. 12 and 13. P. 12 shows my Rivendell, p. 13 a brand-new Singer (not mine, unfortunately). If you overlaid the photos, you'd see that saddle, bars and pedals are in the same spots - the Singer rider has a very similar body shape to mine. Only the top tubes differ. In fact, I would have spec'd the Singer 1 cm taller yet.
Jan Heine, Seattle