I'm for bribing Toku out of war with Zara, if possible. We want him attacking Saladin, or at least Justinian.
No, I calculated Delhi running a single engineer.
That can't be right - you're showing 42 gpp per turn in your spreadsheet. That's 3 specialists. I think we need to plan out another GP pop from Delhi at highest possible GE odds, and run scientists as required. A rep scientist is almost as good as a river village, and certainly as good as a non-river village, IMHO.
UN:
If we don't get an Engineer, we'll obviously rush it as much as possible. The unfortunate thing isn't how long it takes, but that we can't fall back on our Saladin-UN-gift backup plan. We'd need to ensure that Toku is the opponent. Can we estimate their populations with any accuracy from this point? I don't know enough about how growth potential translates into actual city size in AI hands.
Edit: Saladin's kinda on the way to barb continent. One option is to get Toku into a war there later, and help him out a bit?
Settling:
Does it really make that much sense to borrow fish from Sugar? I don't really know the math, but I would figure that the 6fpt ultimately translates into 3 pop regardless (assuming we have enough happy to maintain the pop in one city).
Edit: Actually, forget this. We should settle it.
Deer and cow are definitely ok - not that much pop, but it's something.