Actually, with a harbor they go up to 3 food. But I'd rather spend 60 shields on a granary than 80 on a harbor. Not only that, a Harbor would probably come WAY too late to have a good effect on a settler pump; compared to a granary the harbor's near useless.
My personal plan is to put down a VERY loose city placement, filling in gaps with cheap libraries(but not enough to trigger 100k; I want conquest here. If all goes well I can get my 6th shield in GOTM42 and be done with this run) and just building up military and gold with my initial set of cities, then go all-out with upgrades at some point(might go the horse->knights upgrade route) and storm my nearby opponents, NOT filling in that territory(don't want to trigger another domination win) and then waiting around for the late MA to storm the other continent with Cavalries.
Actually, with a continents map, getting map making early is important if you're gonna get a successful suicide galley, and a successful one on continents is potentially even more powerful than one on archipelago: You meet a ton of civs in one fell swoop, instead of one at a time, and there's a much higher chance that they haven't met before either because of the greater gaps.
Oh, and yes, I did mean 1 NW for the initial move. I get West and East mixed up all the time so that's why you see me saying things like "1 Up-Left" a lot.