So looking at a vaguely sensible alignment, that avoids carving up any of the West Berkshire villages -
http://imgur.com/a/BzfvI
(Satellite imagery courtesy of Google)
60.62km, 728 seconds at 300 km/h (via "new" High Speed diversion)
vs
66.57km, 1192 seconds at 200 km/h (via existing Main Line)
Gives a grand total saving of 464 seconds journey time.
Using HS2 estimates as a benchmark the likely costs for such a scheme would be in the range of £1-1.5bn, never mind new rolling stock etc capable of exploiting the new infrastructure.
Such an alignment would avoid both Didcot and Swindon stations, both of which are quite busy with London commuter traffic. It would also still ultimately feed into a 2 track bottleneck at Wootton Bassett Jn.
***
Bottom line is, freight pathing between Didcot and Swindon really isn't a capacity constraint. Don't forget the existing service level between the points is significantly less than what was extant 10-15 years ago, when not only were there very regular heavy coal trains trundling back and forth between Avonmouth and Didcot Power Station, but the 90mph Oxford to Bristol TM turbos as well.
Even reinstatement of additional loops is probably overkill. If the existing connection from the Down Main to the Down Goods at Highworth was raised up in to the 60-75mph bracket, it would save freights being brought to a crawl to use it - at the moment it's often better to allow the freight to run to Bassett because it will pull a long way ahead while the catching HST does station duties at Swindon.
The key to unlocking a lot of capacity on the GW west of Reading is timetabling. The present clockface 15 minute service doesn't lend itself well to creating additional paths. If, as often seen on Sundays, the South Wales was flighted right behind (or ahead of) the Bristol, you would have 26 minutes free space behind the South Wales between Didcot and Swindon to fill with all manner of new services and freight trains.