you may want to read up on 3rd party ink fading at Wilhelm Research.
A little side note : WIR's papers do seem to apply more on office depot-like inks, the cheapest of the cheap ones... It may well be that the ink4you link above falls into that segment, given their "scanned print" approach for demonstrating ink quality .
If we're talking about more serious ink manufacturers (like Lyson, Cone, Image-Specialists, InkJetFly, etc... their gamut is [a href=\"http://www.ripitgolf.com/ink_comparison.htm]not that bad[/url] btw), they're not in the scope of Wilhelm's tests AFAIK.
There is at least data available at Aardenburg Imaging
, concerning (among others Epson and Canon inks) MIS vs. Epson ink in a R1800 on Epson Luster and Red River UltraProGloss.
Executive summary : the MIS inkset (I've heard it was made by Image Specialists) performs very poorly on the Epson paper, but on the 3rd party paper it ages "only" about 2-3 times faster than OEM ink (much better results than with the Epson paper, which seems to have a specific aging issue with the MIS inkset, or rather vice-versa ).
Mileages vary wildly here, and furthermore it's up to everyone to sum up own conclusions from those raw figures.
And for the validity of feeding printers with 3rd party inks... (cling, the coin) I'll just sum the debate saying that in the particular case of an amateur (not selling prints)
, a CIS (and therefore, often 3rd party ink, even if it is not compulsory) is the only way to afford a desktop printer (the ones like the OP considers, with tiny & expensive cartridges).
For someone willing to invest a bit more, I fully agree that it is much sounder to go towards a pro printer with big
OEM cartridges - and actually, a 3800 is as cheap or cheaper than a 2400/2880 if you consider the ink in it!
Oh yes, and to add a bit to Jonathan Wienke's image, it may be a bit more like cachaÁa in the Maserati's tank, rather than yak urine, in the best cases .