And with Capstone Phase II that would look like this:
In my view, the problem then becomes what happens to all those 1,000 hour wonders when they're halfway between JNU and GST, or perhaps BET and RSH and the magic goes dark.
Carry a handheld (or iPad for that matter) and keep a spare attitude indicator in clear view of the pilot (not over on the right hand side out of the way). 51AK was set up that way and it was great. Problem solved (for the most part). It's also important to train guys how to extricate themselves from those types of situations - even if they're 500hr pilots in 207s. As usual, the solution to these problems are better equipment and better training.
The big thing to keep you from hitting terrain is to be on a pre-defined route at a pre-defined altitude, that's why IFR works. If you can't do that IFR for whatever reason (i.e., limited equipment, limited route structure, etc) you should have a way to do it VFR. In an ideal world, the way to do it would be to build a "company route" that gives you minimum safe altitudes to fly on a particular segment that are not to be deviated from. If possible, make them into pretty maps that the pilots can carry with them in the cockpit to become familiar with the local terrain (not really a problem in BET, but that would be super useful elsewhere, especially for the new guys), and design the routes so that at any point the pilot flies into the clouds, he's able to make a 180° turn safely to fly back to where he came from. Combine this with a training program that emphasizes weather decision making, teaches the 500hr VFR guys how to actually fly in the clouds if they end up in them, and then spend the money to do recurrent training every 6 months with the guys until they are able to upgrade out of the sleds and into Caravans where they actually
will be flying IFR, and a lot of the CFIT problems will likely go away. Additionally, build up the idea that silencing the TAWS or GPWS or TAWS-B or whatever system is installed is a terrible idea. Train that a terrain alert is an emergency condition and they should react accordingly. I have only ever worked one place where we actively trained for avoiding a CFIT and had a specific procedure that we trained and checked on every six months. It was freaking brilliant - "Pitch and Power, Trim and Turn, Avoid the Thread, Protect your Power." I can still remember it and I haven't worked there in a year and a half.
If you want to save lives, give guys the tools they need to not hit the ground. The FARs are wildly inadequate to protect guys in this regime of flying; you need more than the bare minimum. You need better equipment, better training, and institutional policies that take the guesswork out of rock-dodging. I remember plodding up the Casement in flat light and crummier weather than I planned on exploring it in, wondering just exactly how accurate the terrain map was. That shouldn't be the standard - the standard should be a "route" that is religiously used when the weather is lower than some metric. "Oh, yeah, we are always on the company route unless the ceilings are better than 3000', and the vis is greater than 5 miles, then we can go direct as long as we can do it at 2500 AGL, if the ceilings start getting close to us at 1000' AGL, we turn around." Is a lot better than, "well, I dunno, it looks a little iffy. Maybe if I drop down here to 500' I can pick my way through it. I bet it will lighten up as I get closer to Canue Pass, still, it's pretty misty...is that 2 miles? Well, can't turn around now, I'd just have to slog back through all this garbage." Take the guesswork out of it, and the job becomes a lot easier - which is safer.