If I counted right, we're well over the (5/3)(OCN) limit for the map, so palaces are a cool 1,000 shields. That looks about right, judging from the number of columns in the currently-building palace's box.
So an elaborate palace-to-JSBach-to-Smith's plan shouldn't be necessary.
I guess the next step is to figure out some tech costs and try to get a more definite answer than the ballpark 30-40 figure we were given earlier.
Figuring out when we're likely to get the tech will, I assume, help us decide where we're going to actually
do the prebuild.
--
Edit: I did some extra mathematical grunt work with
this formula and came up with some figures. Someone should probably check my work.
If you figure we need education, banking, and economics to get to Smith's Trading Co. (Theology's almost done):
Education has a base cost of 44, banking 52, economics 56.
Our map modifier is 240, for a standard map.
Our AI cost factor is 9.
Assuming that no other civs have learned the tech, and there are four civs remaining after we beat the Persians, we find that the (1 - (N/(7*CL/4))) factor is 1 (since N = 0) and the other constants reduce to 80/3.
Then education has a gold cost of 1173.33, banking 1386.67, and economics 1493.33.
Add them up and you get 4053.33 total gold.
If we put...
200 gold into science a turn after getting Theology, it'll take 21 turns to go this route.
225 gold, 18 turns.
250 gold, 17 turns.
275 gold, 15 turns.
300 gold, 14 turns.
I'm tempted to wait and see how our economy looks in a post-GA Republic before rushing to conclusions.
Edit 2: However did I end up with a number much lower than 35-40 turns?