There are lots of ways to skin this cat. The least expensive would be to not use a pressure tank at all. Let the permeate line from the RO unit drain into anything convenient such as your HLT, carboys, jugs or a polyethylene tank obtained for the purpose. If you want to mount that tank higher and use gravity to fill on brew day, that's fine. For years I brewed in the kitchen but had the RO system in the basement. Over the course of a few days I'd fill a large tank and then pump the water upstairs on brew days with a general purpose pump I bought at Home Despot. I cut a hole in the floor in the cabinet under the kitchen sink and ran a garden hose and an extension cord for the pump through that hole. When I needed water I just put the hose in the vessel I wanted to fill and plugged in the extension cord. Crude, but it worked very nicely.
Now, do these systems deliver the rated amounts? Yes, they do provided the supply pressure is high enough. Essentially RO works by pushing water molecules through holes so tiny that only water molecules will go through. The minimum (theoretical) pressure required to move water through these holes is that sufficient to overcome the osmotic pressure of the dissolved stuff which in your case is a relatively hefty 4.9 psig (because the water is so hard). Beyond that level the flow rate is more or less proportional to the pressure. The manufacturer ought to be able to tell you how much flow you will get with a given supply pressure. Remember that while the permeate flow is relatively low the concentrate flow may be higher by a factor of as much as 3 so that feed flow can be several times permeate flow. This can be significant because of the pressure drop across particulate and carbon block filters which precede the membrane.
Related to this is the question of waste (concentrate). Fancier units are equipped with high pressure pumps, a needle valve which controls concentrate/feed pressure by restricting concentrate flow at the concentrate outlet and, in some cases, an arrangement that allows some of the concentrate to be fed back to the input. Adjusting these controls gives a measure of control over the amount of permeate produced per unit of feed. On my system (no feedback) I can get a liter and a quarter of permeate with about half a liter (it's off the low end of the flow meter scale so it's a guess) of concentrate per minute. Thus my recovery is 1.25/1.75 = 73%. I don't usually run it that way, though. 50% recovery (1 gal down the drain for every gal into the atmospheric tank) is more usual. The GE units I used to use had 25% recovery (3 gal down the drain for every gal of permeate). This is, as you have hinted, only an issue if a whole house system in installed. Whole house systems are way more expensive than the systems you are looking at. Where people are drawing from a lousy lake or briny well they are justified. That's what they are made for. Your concern with recovery was tied to waste in a whole house system. Whole house systems have the controls I have been describing and allow the kinds of recovery (half a gallon wasted for each gallon and a quarter produced) I gave as an example. If you were going to do whole house you wouldn't plumb outdoor bibs, the toilets or the lawn sprinklers to the RO system. You might only plumb drinking (kitchen and bathroom faucets) and, of course, brewing, loads to it. You could continue to operate the dishwasher, clothes washer and bathtub/shower from the softener.
I only mentioned the ROmate tanks because that's what I have and you will be struck with the similarity between them and the pressure tank you already have for the well. I'm sure the people that sell the RO systems you are considering have a line of tanks of convenient size equipped with John Guest connectors. They show one in the video JNye gave the link for. This video will also make clear how the "Y" connection (which is a "T" in the video) is made at the pressure tank.
