If they have an emitmax in the deep violet to uv range then they should be OK. That being said, I have a feeling they are more of just a purple LED. Whether that works or not depends on the resins used on the boards. I really don't know enough about UV development to say much.
I used 100 (99) 3mm types. I wired them in strings of 3, 9 in a row * 11 rows, 1in spacing. Driven by 12V regulated, with a couple of series power diodes to lower the voltage to ~10.8V. No dropper resistors.
I have to stand the target about 15" off the LEDs to get the spots to merge. I can do an A4 size PCB. The exposure time is 6 mins. The illumination is a bit patchy because the LEDs aren't exactly perpendicular so the projected dot pattern isn't entirely regular.
I guess you'd have a 5 * 4 matrix. If you spaced them on a 2" grid that would give you 10 * 8 inches. The collimation will not be as good as with many smaller ones, but this is only really important for extermely fine detail. You could run 2's or 4's or 5's in series, or of course singletons. The wavelength is right. Other than that there are no guarantees, it's going to be experimental to a degree no matter what you choose.
I've dealt with Asia Engineer on many occasions, they've always been reliable. Assembled quality (soldering) of some of their meters is a bit untidy, and delivery often takes over a week, but other than that, no complaints where LEDs, displays and other components are concerned.