UPDATE 8/8/23:
So I wasn't happy with how the photo looked on the Rag Photographique. Looked better somehow on the Arches 88. Interesting how a photo can look different on papers which are so similar. Maybe it's the different color casts of the papers — Rag Photographique toward yellow and Arches 88 toward pink — that affected the look. Or the texture which is a little different for each paper.
Anyway, I decided to try again to get a print without lines (banding) on the Arches 88.
I replaced the two ink cartridges that were low, so I could do another power clean. Those were the cyan and vivid magenta cartridges. I had gotten "low" warnings for them, but I usually wait until I really have to replace before replacing. But the manual said cartridges have to be at least half full for power cleaning, so I replaced.
Did the power cleaning.
Printed the image on Arches 88. No banding!
Settings in Photoshop print settings window: 2880 dpi; high speed unchecked; media type - ultra premium presentation matte; paper width 4; platen gap auto; Canson ICC profile.
I made 6 prints. I fooled around with curves a little when printing, that's why I made so many prints. And once I got one I liked, I made an extra copy.
On the sixth print, I tried printing with high speed checked, to see if that caused the lines to return. It didn't.
I also wanted to see if non-high speed affected darkness of prints, because a print from a couple of weeks ago I made from same file — printed in high speed— looked lighter than when I printed it today in non-high speed. I thought maybe if not in high speed, the ink goes on heavier? But the darkness today in that file was the same in high-speed or non-high speed.
I wonder if the lack of banding today on the Arches 88 was due to the fact I put in new cyan and vivid magenta cartridges. Epson says in manual that low ink cartridges can cause banding. The banding was in area where there would be a lot of ink (an almost solid black area and a dark blue area). Maybe not enough ink coming out? But the cartridges weren't empty, and I wouldn't think they were so low that not enough ink would come out.
Still weird that, last week, I could get prints without banding on the Rag Photographique but not on the Arches 88. The Arches 88 must be more sensitive to slightly clogged heads or low ink cartridges. Perplexing.
As a side note, I found out about how much ink is used on each 8.5 x 11 print, so I was able to calculate how much I spent on ink and paper for each print. Comes to about $4 per print. Which means I spent a fortune on this project! I'm not even going to say how many prints I made in the past few weeks while dealing with this banding issue! It's ridiculous.
But after I calmed down, I told myself I should know better next time how to deal with banding and won't need to make so many prints. I figure if I allow myself 10 prints per image ($40) to get the print right, that's not too bad. It would cost that much to get a really good inkjet print made at a lab, and I wouldn't have the opportunity to revise it ten times.
Another side note: I remembered today to use blower on the paper before I put in the printer. I had forgotten to do that before — which means little dust spots appeared in a number of prints and made them unusable. There was either a glob of ink where the dust was or the dust fell off and left a white speck.
Maybe these long posts will help others someday who are researching Arches 88 and banding, or struggling with home printing in general : )