having dove the evolution for a year, I sold it because my wife's had lots of electronics problems, and that was with no tinkering or modifications. To achieve the level of true electronic redundancy I was comfortable with, to allow for a variety of tank sizes available when traveling, any modifications to the harness, removing the cowling and replacing with a travel frame... etc, it got to the point where purchasing a copis meg and practically starting from scratch became the easier and less expensive option.
If I were to get the evolution today, I'd at least add a forth cell monitor (a shearwater pursuit), and make it so that if the vision controller failed on a remote trip I'd be able to abort the dive and once back on the surface plug the shearwater into three cells and run the rig manually for the rest of the trip. A significant failure on a vision invariably means a trip back to the factory, and that can mean not just a lost dive on a trip, but an entire lost trip (for us it was two major trips in a row that were nearly ruined) if you don't have a back up for three cells. One of the main reason's I have reverted to a manual system is that they are more field serviceable and a lot more modular. the new hybird HH give you more options in case of a fialure if you really think you need a set point controller...You get two independent handsets and the ability to switch back and forth from e-mCCR.
The vision was a dream to dive, but I really had to work hard to keep paying enough attention to catch that one in a million failure in time.
If you are looking for modularity, the evolution is one of the least modular of rebreathers I've looked at. If you like it the way it is, it's has the best eCCR controller on the market, and the temp stick is a very nice tool... IMHO.