Although it appears that nirpat89 has settled on the P400, I wanted to better understand these Red River numbers.
Taking just the RR PK numbers, the P800 prints an 8x10 page for 68 ¢ and the P600 does it for $1.02, so that's 50% more expensive per page on the P600. But this is dependent on the relative price you're paying for the two different types of cartridges. I went to check the Red River cartridge costs and they no longer sell Epson cartridges, and so we can't check what prices are behind their cost per page numbers any more. But since they do provide the number of cartridges used, we can construct our own.
And I found that it varies a lot from country to country, and perhaps from retailer to retailer, varying from a few samples. All were more favorable than the RR numbers, varying from the P600 being 55% more expensive (Epson NZ) to 66% more expensive per page for that standard page that RR used for testing. At current B&H prices the P600 would be 58% more expensive. So the payback period for purchasing the P800 is going to vary, but seems to be shorter most places than if based on those RR numbers.
There are a couple of other factors that are relevant. If we stick with those B&H numbers for the moment, a single P600 cartridge is around $32 which implies $1.24 per ml for the 25.9ml of ink. The P800 cart is $55 or 69 ¢ per ml for the 80ml. So the P600 ink is 80% more expensive per ml. So why is the P600 only 58% more expensive to print on?
One possible explanation is that the P800 does more head cleans than the P600, or uses more ink in doing them, which I doubt. There's a question of how RR ran their tests, and whether ran the P800 in ways that generated more cleans. Hard to know.
The other distinct possibility is that you use a smaller proportion of the ink in the P800 carts before they register as empty compared to the P600. I think that's the more likely explanation.
I agree with Bob that with the P800 you have to work harder to get the maximum out of the cartridge. By which I mean that if you change it when the printer first insists you do, because it refuses to do a nozzle check or black ink swap until you replace it, then you're not getting the max out of the cart. Usually if you replace the apparently empty cart with a full one, do the nozzle check and then put the empty one back in, you can keep printing for a while longer. Given the nature of the testing that RR presumably did, I bet they didn't wring as much out of their P800 carts as they might have. Nor would they have tried to harvest or sell the considerable quantity ink that remains in a P800 when it's as empty as you can make it. This would widen the gap between the cost of printing with the two printers and shorten the payback period, as I don't believe that the P600 carts have as much residual ink when empty.
The other factor that is relevant is how much is the real price differential between the two printers that you'd looking to offset? Again using B&H prices as a reference point, the P600 costs $796 and a set of carts costs $273.50, so the implied cost for the bare printer is $372.50. The cost of a P800 is $1,195 and the inks are $494.55, but you only get 64ml starter carts, which I value at $395.64, making the cost of the bare printer $799.36. So the price difference for the bare printer, which is what you have to offset, is $276.86, not the $399 difference in the sticker price.
Moreover, for those in the US, there is a $300 mail-in rebate on the P800 and only $150 on the P600, so after rebate the implied difference in bare printer price is only $126.86. That's a lot less than 2,000 standard 8x10 pages required to offset the additional cost of the printer, like only 326 for those eligible for the rebate. Personally I regard that as a no-brainer, but YMMV. I think payback period would be lower still if you did as Bob and I suggest and really use each cart until you asolutely have to replace it, and harvested or sold the unused ink. (I'm still ignoring the ink used in the initial fill, as I assume that it's a similar quantity of ink in both printers, although it's cheaper ink per ml on the P800.)
There was an interesting article on the P600 vs P800 issue on TOP. In his featured comment MHMG makes the same point as Frodo - that you need to print enough to make up the difference. I wonder if he falls into the trap of not discounting for the value of the additional ink that you're buying with the printer. He doesn't mention it specifically, although his statement that "The lower ink costs of the P800 versus the P600 only begin to bear economic fruit after you've printed several hundred 8x10 equivalent size prints or more" is in line with the above figures. Of course if you're printing 13"x19" or 17"x25" then the number of such pages falls even further.
http://theonlinephotographer.typepad.com/the_online_photographer/2016/09/p800-vs-p600-the-epson-value-equation.html