Whew, a blizzard of information and questions; appropriate for the season, I suppose.
I've never had a 4990, although I know they work fine for this purpose. My V700 has two lenses: one higher-res that focuses off the glass, and one lower-res that focuses on the glass. The higher-res lens can only accommodate up to about 5x7. I scan film with the holders, and would make one if I didn't have them. I shoot plates by placing them on the glass, emulsion side down, propped on four coins at the corners (US dimes, which are of about the right thickness), which works well for me. You can place them directly on the platen, but of course you risk Newton's rings appearing in your scans, which I assume you already know from the mention of ANR glass. In my experience, sometimes it happens and sometimes it doesn't.
Your suspicions about the ASA calculations are correct. You do the best you can with your meter, then add stops until you get there. This turns out not to be too bad in practice because a) it doesn't take long until you get a sense of the correct speeds for common scenes, and b) you can use one of the light meter apps on your phone, some of which go down pretty far in ISO setting. I often use "Pocket Light Meter" for the iPhone, and it works well enough for me in many circumstances. I used to use my old Gossen Luna Pro fairly regularly until it fell off it's hook and landed in such a way as to destroy its incident dome and crack the surface of the sensor. May replace it, but haven't got around to it yet.
Jason Lane's dry plates are good. I chatted with him at a local photo show last year -- very nice guy -- and bought a box of whole plate-sized plates to try out, which saves me from having to mess with cutting a new glass size, but I haven't refurbished the whole plate camera I have yet to try them. In general, I tend to prefer making and shooting my own plates, but that's just me, has nothing to do with his plates. It's not difficult, but does require an investment in accoutrements, and an appropriate (darkroom conditions, fortunately under red light) place to work. But developing plates, and paper, is great, because you can do it by inspection once you have a place set up for that.
One note about both paper negatives and dry plates: as I'm sure you aware, they are not panchromatic. What the paper is sensitive to depends on the paper: may be blue only, will be more like orthochromatic if multigrade. The main point here is that meters are usually calibrated for panchromatic response, and you cannot really trust them with "colorblind" plates and some papers in anything but sunlight. Not that you can't use them, but that you really have to apply a compensation factor to account for the composition of the light. If it's full-spectrum, you're likely good to go. If it isn't...well, you're going to be adding stops, and likely using your best guess to do so. So low-light, or indoors under tungsten or LED, etc., will mean you're in reciprocity-failure territory and will just have to test until you learn how your material responds. It's fun to play with, but be prepared for it, and not too disappointed if your early attempts aren't perfect. If you want to do portraits, you'll need either a LOT of artificial light, or stick to rather direct window light! I'd recommend architecture or landscapes to start, and still-life indoors next, before messing with portraits in this case. Or just use yourself as a subject until you get what you want; it gets boring staring at yourself, but removes a lot of pressure from the learning process. DAMHINT. (Don't Ask Me How I kNow This)
The whole process is fun, though, so I suggest you just jump in and give it a try before deciding which direction to go next. I spent too much time trying to overthink things and eventually found that just doing it was more rewarding and fun and gave me better data to help me figure out where to go next.