It’s not clear whether Harold Wilson ever actually said that a week was a long time in politics, but then uncertainty about time is one of the defining themes of the past century or so, since Einstein messed everything up with his general theory of relativity.

Political time is especially fluid: everyone likes the similarly semi-apocryphal line attributed to Zhou Enlai, when he was asked in the early 1970s about the impact of the French Revolution, and replied that it was too early to tell. In fact, Zhou was talking about les evenements de mai 1968, but the wrong end of the stick is actually the better measure. There’s never a suitable time to make a definitive judgment on whether something has been a political success. Historians, after all, make their careers out of disagreements on the import of events that were stale news before the birth of Christ.

All the same, two decades seems a reasonable interval at which to take stock, as the Scottish Affairs Committee was doing this week in a report to mark 20 years since devolution came into effect. Its conclusions were that the relationship between the governments at Holyrood and Westminster has “deteriorated at a time when goodwill and co-operation are needed most”. That was the least of it, actually; phrases like “mutual distrust”, “fractious political landscape”, “renewed strain” and “political stalemate” are all bandied around, and that’s just in the first two paragraphs of the summary.

READ MORE: Marianne Taylor: Tory confessions highlight hypocrisy of our drugs laws

Which tells us that “the settled will of the Scottish people” – as John Smith described devolution at a Labour Party conference 25 years ago – turns out not have settled all that much. Indeed, it’s hardly even managed to settle down.

You may have deduced as much for yourself, having noticed that support for the SNP and independence (with the caveat that they’re not always the same thing), decidedly minority stances before the 1990s, are now mainstream, if not quite majority, positions. So much for the notion voiced, particularly by pro-devolution Labour MPs at the time, that a Scottish Assembly (as it was then still called) would “kill Nationalism off”. Devolution, which ought to mean “rolling down”, has in effect rolled over the previous political assumptions of all sides.

To say that isn’t to make any judgment for or against the devolution settlement that we ended up getting; it’s merely to point out that what has happened since is not what was generally anticipated before its introduction. But that is true of most constitutional changes. For example, the disproportionate majority of Nationalist MPs, compared with the overall popular vote, is in itself an indication that no structural tinkering ever provides a perfect reflection of the electorate’s priorities.

The two things – the rise of the SNP and the Brexit vote – that the committee’s report cites as central to the current tension around devolution, and which are certainly the biggest and most divisive issues around, didn’t back then feature much in most people’s political concerns, unless they were especially committed to Nationalist causes or the Bruges Group. By the time that they did come to the centre of political debate, the solution in both instances was to hold a referendum that would settle the issue for once and for all. How’s that working out, do you reckon?

There’s another aspect of devolution that’s now obvious, but wasn’t much anticipated, which is that the stance of the national wings of the UK-wide parties tends to diverge – just this week, Scottish Labour has come out unambiguously in favour of a second referendum on Brexit, while the English, or at any rate Westminster, incarnation of the party continues to resist it. And whether or not Ruth Davidson can work with Boris Johnson, if he were to become the Tory leader and Prime Minister, there’s no shortage of areas in which the Scottish Conservatives will have a different emphasis, if not an outright different policy, from the UK party.

These differences need not in themselves be a problem; you may even see them as a virtue. Most of us, even the small minority who think that the very existence of the Holyrood parliament is a mistake, are in favour of devolution, in the sense that we think decisions are best made nearest to the people that they effect. Even advocates of that most ambitiously centralising project, the EU, express theoretical approval of that kind of devolution. But the practicalities are the tricky bit.

READ MORE: Ruth Davidson backs Home Secretary for Tory leadership

No-one doubts that the Scottish Affairs Committee is correct to conclude that there are serious tensions in the existing settlement. Its proposal for improving matters – to scrap the Scottish Office and introduce a UK-wide department for handling constitutional affairs and liaising with the devolved governments – doesn’t, on the face of it, look all that promising as a solution.

It’s impossible, for a start, for any such bureaucratic change to do anything about nationalism or Brexit. On the latter, nothing can alter the fact that different areas of the country voted in contradictory ways. And the SNP, perfectly naturally, are always going to think that devolution is unsatisfactory when measured against independence.

But the deeper reason tension is bound to continue is that devolution hasn’t been extended to 84 per cent of the UK’s population (with the exception of the very limited powers of directly elected mayors in London and some other cities). There’s precious little evidence that the English regions want it – some areas expressly rejected it – and, worse, quite a lot to suggest they resent its existence elsewhere. They think they’re getting a bad deal, but don’t want to have a similar set-up.

That’s an insoluble problem, whether you’re nationalist or a unionist, because England is so dominant in population terms, but doesn’t lend itself naturally to federalisation. Bar a few places (Yorkshire, perhaps, or Cornwall) there aren’t natural geographical candidates for devolved government. There’s a case that Scotland has more in common with some English regions than either has with London, for example, or, conversely, for arguing that London’s dominant political preferences (Remain, centre-left, environmentalist) are more like Scotland’s than the rest of England’s.

Devolution was a revolution not, as it was painted, a natural evolution. Like the French Revolution, it may still be too soon to say whether it’s been a success, but the one thing it clearly isn’t is a perfect solution.