Ted, here is my interpretation of what would happen:
1. You do a 50% water change with tap water, and KH doubles to 10 dH. The added alkalinity and the addition of low CO2 tap water initially drives your pH up. The additional carbonate buffering capacity would make it harder to drive down your pH back down...you would have to add a lot more CO2 than at a KH of 5 dH. Assuming your CO2 diffuser is adequate and you don't start outgassing CO2 through the water surface as the the concentration in the tank climbs, you will eventually reach the controller pH of 6.8. The dissolved CO2 in the tank will be high (I don't have my chart handy for the exact amount) and would/could stress or kill your fish.
2. You do a 50% change with RO water, and KH drops to 2.5 dH. Initially, the drop in KH should cause your pH to also drop, but you are adding low CO2 RO water, which should cause the pH to rise. Will the two processes balance each other?...Don't know; it's hard to say. Your carbonate (and total) buffering capacity is now half of what it was, so CO2 injection will alter your pH relatively quickly. Depending on how far and how fast the pH changes, this could stress your fish and lead to disease or kill them outright. Dissolved CO2 in the tank will be sub-optimal for plants so they will suffer.
Given these two scenarios, I would add carbonate and total hardness to the RO water to get the desired value of 5 deg. KH and 5 deg. GH **before** I added the water to the tank. That way there is no "error."
BTW, why don't you set up these two scenarios as "experiments" in a bucket? I would be very interested in your findings!!!
29 gal planted tank, 30" FW Aqualights w/ 1x 65W 6700K CF & 1x 55W 9325K CF, 20-lb CO2 w/ Milwaukee SMS122 controller, reverse-flow UGF (bah!), gravel substrate (also bah!), 4 Praecox, 3 Ottocats, 1 SAE, 1 Bristlenose cat, 2-3 C. japonica, 5 Red Cherry Shrimp, & too many detestable pond snails!!!