Great Western Coffee Shop

I notice that Real Time Trains had serious amounts of red figures for services on the Mule yesterday (12/11/19). Although at least one of the services had a train fault, I guess the vast majority of delays can be attributed to congestion caused by the number of services using the single track. This, in turn, was due to the closure of the WR main line at Cogload Junction. I guess, therefore, that any delay attribution will be down to Network Rail?

Over the years, this must amount to £millions. I appreciate that political decisions are nearly always short term, but would it not be cheaper in the long run to reinstate several sections of double track on this diversionary route?

I have always reckoned that the real reason for doing delay attribution with real money is to make that happen. If the costs of not doing something pile up on the same desk as the bills for the related project to fix the issue, the decision should be easier and have a lot less politics involved in it. Of course how the accounting is actually done between cost centres inside NR is another question. You also have to reckon with the natural tendency to put off capital spending, even if justified, because something else is more urgent and funds are finite. And I suspect that the sums that actually get exchanged are only indirectly related to the real costs, which might blunt the effect.

Not necessarily, the GWR diversions have been diagrammed to fit around the existing service and have a very generous running time with plenty of recovery so any delay probably won't directly be attributed to the Cogload works.
 
You may find it all started with the train fault, or the flooding near London, or someone sneezed in Inverness that caused Scotrail to be 1 minute late, which caused Crosscountry to be 2 minutes late etc.  Which may sound daft but several times a  delay on a branch line in Devon has caused delay attribution from it to spread as far as Scotland.

So, you spend a fortune on double tracking a line that works perfectly well as a single line, just for one or two weeks a year of diversions that also work reasonably, but not perfectly well.........

I would rather spend my money on more important capacity constraints such as Bradford Junction to Thingley Junction.

I certainly agree that Bradford Junction to Thigley Junction needs looking at, but punctuality on the ex-LSWR route seems to have plummeted recently, even when nothing other than SWR services have been running.