In which case, the Scottish Government is not really doing its job properly is it?
The job of government is to find what is generally politically popular and sound and try to enact that in ways that offer good value for money.
The fact that money is being spent on projects with demonstrably awful value for money like the southern end of the Borders Railway and the A9 dualling means that Scots will collectively all be poorer in the future than they would have been if the money had been prioritised according to more shrewd assessment of the returns, such as building new stations, making targeted safety improvements and doing CCTV schemes on major roads, bus priority schemes, faster mobile data and so on and so on. Single carriageway improvements with increased overtaking opportunities and safer junctions can bring almost all of the benefits that the A9 dualling scheme will (especially with its all-grade-separated junctions) at much lower cost.
You can of course make the case that there is a great social benefit that is not being captured by the traditional methods - but there is a reason it isn't captured, and that's because it's intangible and difficult to measure. I can see the case for social benefits that result from a railway for Galashiels. I cannot see any that result from driving an unnecessary dual carriageway through the Highlands.
I'm well aware that smaller, targeted interventions can easily have much higher BCRs. Very small schemes can have BCRs in the double digits. My belief is though that the idea that spending £3 billion on all of these things instead would be massively, enormously better than spending it on one
grand projet might not be quite right. If you went about improving each and every junction on the A9, you would end up encouraging more traffic to use it without the main road being upgraded to cope. Each project individually might have a high BCR, but a package of every single one of them combined might not.
What would you do with the full £3bn for the A9 dualling anyway? As far as I can tell, the pipeline for other plausible infrastructure projects with good BCRs in Scotland couldn't actually use all of it. How many stations are worth building and would require Scottish Government funding? Junction improvements are likewise great, but how many of them really are there? Laurencekirk? Sure, but that project is in the pipeline anyway. Nice wee things like the Pulpit Rock works or the Crianlarich Bypass which get rid of major A82 pinch-points have already been done, leaving only big and costly things like the whole Tarbet-Inverarnan upgrade. You'll always be limited by the law of diminishing returns, and the relatively low BCR of the big schemes is because they're plonked quite firmly at the diminished end of that scale. It would obviously be great if there were £3bn worth of small schemes with high BCRs but I really don't see how that would happen.
Investing in better mobile data is a great idea, except for how it's privatised and Holyrood has minimal control over it. I can't see a big cash injection into the major private mobile companies having that much of a long term effect. If they don't want to bother building 4G towers in the Highlands now, why would they bother with 5G, 6G in future? If the government is always the one to cough up, then why bother doing private investment? I don't oppose the idea in concept, but it would require changes to the mobile networks which are well beyond the purview of Holyrood. A Network Rail-style state-owned-but-run-privately benevolent infrastructure-only monopoly where the other networks become MVNOs would be great and would make government investment much more practical.
If you run out of good things to spend capital investment on, then you obviously can look at other things as well. But, there's only so many capital investments which make sense. Housing might be the default demand but it can and should be the case that it doesn't need capital investment like roads do. With the end of right-to-buy councils should now be able to borrow commercially to build housing, using the guaranteed income from council tenants and the scale and efficiency they can bring to construction and maintenance. There shouldn't be many schemes which could only go ahead with central government capital investment, as then it must be the case that the rental income can't pay for them. You don't want councils building homes which they couldn't then sell profitably at market rates.