The problem with WAT-EXD is that whilst there's almost certainly a good business case for electrifying to Salisbury, but the rest of the route up to Exeter is more of a problem. It's unlikely to be worth electrifying without redoubling, and once you add all the costs together, I suspect the BCR falls through the floor. So, if you wanted to free up the copious quantity of Express Sprinters currently running the service, you'd need to solve those problem.
I think I've covered this before, but basically you choose one of:
1. Spend a fat wad of cash on full redoubling and electrifying. Apply dual-voltage 444-alikes
(Oh look, it's the 444-alikes again. At least this time it's for SWT though!)
2. Split the service in half. Make WAT-SAL into EMU territory and run EXD-SAL (or maybe even EXD-SOU) with much fewer DMUs. You could still run the odd few of EXD-WAT through services in the peaks.
3. Come up with some kind of bi-modal EDMU design and use that.
4. Get some different DMUs for the service, remembering that the capacity of a 10x 158/9 is needed on peak trains and platform lengths won't allow for longer trains. In other words, Voyagers aren't going to help you here.
5. Accept that Express Sprinters are stuck here for the foreseeable and look elsewhere for your DMU cascade needs.
Unless one of you clever folks can come up with something I've not thought of.