In their summer 2014 investigation, Network Rail estimated the cost of reinstatement of the North Devon via Okehampton line as £814 million for a relatively low specification, primarily local, railway. By way of comparison, their estimates for tunnelling interventions to improve the existing South Devon line were between £1300 million and £2900 million. More recently they have put forward the interesting notion of a causeway-borne railway 25 m from the existing South Devon line no doubt cheaper at around £500 million, but in my view almost certain to be operationally non-viable at least as frequently as the existing Dawlish line. See
http://www.bbc.co.uk/news/uk-england-devon-37639243. Recent evidence suggests that Network Rail continue to be somewhat optimistic in their planning and forecasting. The cost of an alternative to the Meldon viaduct on a reopened North Devon line, if the existing viaduct cannot be revamped, would pale in significance against such large overall budgets. In addition, reopening via Okehampton has the considerable advantage of creating a new link to areas of the country which currently are not served by rail and also addressing the economic deprivation of North Cornwall. Added to the potential benefits of Exeter to Okehampton commuting and relieving the congested road network to Tavistock, I still find myself much more positive about the North Devon reopening.
If NR are being overly optimistic about costs and forecasts on all projects by the same margins (I.e. everything is 10% more) then it doest matter if NR say it's going to cost £81 vs £130 or £1.8bn vs £13bn the comparable costs would be the same. It is only if they under estimate in one area and not another that things would be hard to compare the costs of the various projects.
Yes, under estimating doesn't help with the cost benefit analysis.
One question I have is, does the reopening allow for the section to Tavistock or is it assumed that section is open? As if it is open, and it assumed that it wasn't, then there is likely to some cost saving. Also, as I highlighted previously the cost of running through trains is not that more than running the two branches and so the benefits are increased.
It is likely that the reopening would mean that other schemes would be more likely to happen, for instance the redoubling of the WofE line and arguably the Dawlish Avoiding Line (DAL) as well. In that with more passengers heading west of Yeovil the need for a more frequent service increases, potentially to the point that you have 2tph all the way along the WofE line which both extend to Plymouth. Likewise, with an increase in passengers between Exeter and Plymouth the justification for a faster route increases.
Conversely, the likelihood of a more frequent service along the DAL lays with a bigger demand for services "up country" (I.e. London and/or Birmingham) where there are signification capacity issues. At least on the SWML there are plans, in the form of Crossrail 2, for capacity enhancements which could allow 3tph (semi fast services) to Salisbury with the possibility of at least some of those being able to extend to Exeter. The GWML would require more tracks into London or the extension of the Bristol services. Whilst the route through Birmingham New Street, although will benefit from HS2 also has a lot of competing routes which would also benefit from extra services.
Also as I have indicated before there is at least the scope for the route through Okehampton to attract passengers from a wider area. In doing so it means that they fill up services which are either emptying out as they have got beyond the main capacity issues with London (I.e. Few people will travel from Waterloo to Plymouth, so the extra passengers will be joining at Woking, Basingstoke and Salisbury where trains tend to be that bit less busy) or are new (I.e. the bit between Exeter and Plymouth).
It should also be noted that even when competing against a service frequency of 2tph), depending on the timetable it could still be quicker to get a slower train than a faster one. For instance if the slow train leaves 10 minutes after one fast train but still gets to where you want to go 5 minutes before the next fast service then people are still going to use it.
The other thing to remember is that not all Exeter to Plymouth passengers start at St. Davids, as such there could be significant numbers of people who would use a direct service from Central over having to change, even if it was (even allowing for a change) slightly slower. This would be especially true as it could be that you then effectively have the choice of 2tph (one where you change and one where you don't) even if they were broadly at the same segment of any hour (I.e. 10 minutes apart).