get the ICC profile from Pictorico
I don't see how that would solve a banding problem. I think maybe the Epson rep misunderstood the problem and was thinking of posterization, or they simply don't understand very well how this problem works.
The problem with the banding you mention (I think I recognize it) is a fundamental alignment issue that's basically impossible to solve. You can work around it by using more color channels at the same time (c.f. piezography), which will effectively 'smudge away' the problem. This is because the inkjet nozzles are arranged in a distinct, geometrical pattern and evenness of tone is achieved by jetting quasi-random dot patterns from the nozzles of several channels. If you print on regular paper, ink bleed is crucial in making tone appear evenly. OHP film doesn't have as much ink bleed (the droplets basically stay right where they land, with sharp edges).
How are you printing your digital negatives - i.e. from which app(s) and with which settings? I generally use QTR which allows you to make custom curves that use several channels at the same time. Using e.g. Mk, LK, and Y at the same time may help.
There's a bit of a tradeoff when it comes to using multiple inkjet channels:
* The more channels you use, the more volume of ink (=water, mostly) is deposited onto the film. OHP film will only take so much ink before it starts to pool and dry up in ugly patterns (or not at all), so there's in practice a limit to how much optical density you can make with channels that give relatively little blocking power (e.g. LLK, C, M).
* The fewer channels you use, the coarser the tonality and the bigger the problems with banding/striping etc. will be.
Try to strike the best compromise by combining as many channels together as feasible without running into ink pooling problems. This will be easier for processes that require not too high density; in this case you're in luck with your Pt/Pd prints as they only require something like 1.8logD (off the top of my head). Things get much trickier with salt e.g. prints that like 2.2logD, which makes it harder to get smooth tonal rendition.
I feel that getting even tonality is the one, major tricky aspect of inkjet digital negatives, and to be frank, I find the problem such a nuisance that I just cannot actually enjoy printmaking from inkjet negatives. It always ends up as a disappointment, at least in my hands and within my expectations.
Here's a snippet from a carbon transfer step tablet test; it's an imperfect transfer onto glass (note the broken line on the right), but it demonstrates the problem IMO.
The Stouffer step tablet to the left is a silver gelatin negative, basically. The 3-column step tablet to the right is a DIY one inkjet printed on my Epson 3880 using the MK, LK and Y channels (it's not linearized). You can already see the coarseness in the mid-grey steps in the bottom half of the rightmost column.
Here's a closer look at that bit:
Note that this also has banding and that the rendering is rather coarse. By comparison, this is a section from the Stouffer tablet at the same magnification:
Now, I know I could get matters to improve at least a little by buying a new printer (€1k and upwards), convert it to piezography (IDK what that costs now - if it's technically possible to begin with on these newer printers; maybe starting at €700 or so?). But the improvement would still be marginal IMO.
Sorry about the rant. I can only say I sympathize with your problem. There are probably things you can do to improve matters somewhat. The use of a different ICC profile likely is not going to solve this problem.