My proposal would be a new chord from Hele to Pinhoe alongside the M5. This, in conjunction with a reopened Okehampton route would provide a route without reversal and give GW services access to Exeter St Davids avoiding Cowley Bridge if flooding occurs there. You'd then need problems at both Cowley and Dawlish to cut off Plymouth and Cornwall and, if either flooding or sea wall difficulties occur, a service with only moderate delays would be maintained for Plymouth and Cornwall.
About 10 km alongside the motorway all the way. As there are no road junctions along this section, roadworks should be minimal, perhaps largely widening bridges and earthworks. Would have to cross the M5 at some point to swing round towards the west and join the LSWR just before Pinhoe. The existing motorway overbridge there is only wide enough for a single track.
I would think to solve the Exe flooding issue it would be a lot less expensive to raise the existing route by two or three metres onto a new properly drained new structure for two or three km through the Cowley Bridge area. A problem with that is the A377 overbridge which would have to be renewed with a small road diversion for clearance, although that could allow a remodelling of the junction layout with an additional independent track or two for the Barnstaple/Okehampton route.
The Okehampton route could be largely single with dynamic loops at Crediton to Yeoford, Okehampton and Tavistock. The normal service could be provided by an extension of Cardiff to Taunton services serving the proposed reopened stations at Cullompton and Wellington via the Hele chord thus serving the very popular shopping / commuter destination of Exeter Central and then continuing via Crediton to Plymouth.
Would agree that if reopened on its own merits, a fast single line with dynamic loops would be the most practical and economic track configuration for the Okehampton route, although I think the most appropriate train service would be an extension of the Waterloo - Exeter trains to Plymouth with a limited number of stops. In addition to connecting stations en route to both Plymouth and Exeter, Plymouth would regain through service to Honiton, Axminster, Yeovil and Salisbury, and possibly Exeter Airport. The route wouldn't be competitive to London perhaps, but could nevertheless provide some very useful local and regional connectivity.
I am still convinced a DAL (together with retaining the existing coast route for locals) offers the best package of benefits for developing longer distance services to Plymouth and into Cornwall whilst preserving the local connectivity and cost effectiveness of operations along the route at attractive frequencies and journey times. However I am also keen to see the Okehampton route reinstated in some form at some time in the future as a regional link. This might be best accomplished by progressive development of the 2 branches at either end, as has been proposed already.