Blub, I can't say much about the differences between 3.13 and 3.17. In general, the maintenance cost for overseas cities in BTS is crazy outrageous. We all know how to manage WW, so I don't think that's a huge issue. Fight short, decisive wars and WW doesn't become a problem.
I'm fairly certain that there's not enough time to get a second city productive and still get the Oracle built there. At any rate, the capitol should be dedicated to priest wonders, with other wonders going elsewhere. This supports an early Stonehenge strategy, and the power of settled priests in the capitol is formidable.
I would (and often do) target Stonehenge, Oracle, Chichen, Shwedagon, Angkor, basically all the prophet producing wonders for the capitol, and relegate the others to other cities. A specialized GE city could build the Pyramids and the HG, as well as the Kremlin if for some reason we're playing that long. I'd probably either stick the Colossus in the Capitol or the engineering city. I usually use a third city to build the Parthenon and Glib, and any other wonders. This makes the appearance of a GS lsss certain, but it's a risk I'm personally willing to run in my own games, and one that can be mitigated to a large extent both by the scientists given by the Glib and by running additional scientists.