There's a good deal of variation. Without digging into how that figure was produced, it could be worthwhile to mention that the production process Jay Ruskey says he's interested in cuts a lot of water out of the process. Once coffee cherries are picked, getting down to the seed inside (aka "coffee bean") can happen one of roughly three ways: "natural process", letting the coffee fruit dry out like a raisin on the seed; soaking/fermentation, letting the coffee soak in water for 24-48 hours once (or often twice in Kenya); mechanical demucilagation, separating the seeds from the rest with a machine. The demucilager is common in Costa Rica. I can dig up some more material later, but the government in Costa Rica is strict about water waste, essentially requiring farmers use demucilagers to process their coffee. I don't know how much this would affect the water involved in coffee production, but I would guess it would drop the amount necessary for production in cases like Ruskey's.
Coffee seems to be more comfortable in warmer and drier locales than tea though, and very high lands. Furthermore, I don't know about tea but coffee can be shade-grown in polycultures e.g. in the shade of fruit trees.
> Coffee requires almost 10 times as much water [than tea], using 1,056 gallons of water per gallon of brewed coffee.
That's a lot.