I accept that, but rewilding is just an extension of the same idea. It is still about 'recreating' - name a single 'rewilding' effort that does not involve a reintroduction (i.e. an intervention), or includes a full complement of species. Come to think of it, name a single 'rewilding' that has been successful. Yellowstone does not work within the context of its surroundings (the wolves are managed).
On what basis do you see Yellowstone as a failure? I haven't been there myself, but by most accounts the wolf reintroduction had been pretty successful. Wolf management seems to be minimal, and is mainly killing of animals outside the park. Inside the park, numbers are self-regulating. The culling is regrettable, but hardly invalidates the reintroduction. The key thing here is that culling is aimed at reducing direct wildlife/human conflict outside the protected area, rather than at micro-managing the ecology within the park - this is not at all comparable to active habitat management (e.g. tree/scrub clearance) on nature reserves.
Similarly, the Scottish and Welsh beaver reintroduction projects are likely to result in some management of animals/dams being required in particular problem areas, but this is quite different to saying that the population 'needs' to be managed to benefit the ecology (in the way that a meadow 'needs' to be managed).
I have highlighted the key words here. Not only are they extremely vague, but they also represent a 'target' to reach, i.e. a state to achieve through action (not inaction). To recognise this, all we have to do is ask ourselves when a project has succeeded. How would we know? When would we stop intervening? When do we stop introducing species, or managing them? Probably never (even the 'wild' Bison of Europe are managed, culled, fed, treated by vets, never mind the domestic horses in the Dutch 'rewilding' experiment).
Personally, I see the main requirements for intervening as being (1) to reintroduce (locally) extinct species, especially the keystone ones such as wolves and beavers which are unable to recolonise naturally (unlike on the continent), and (2) to control non-native invasives (e.g. rhododendron) which have not co-evolved with native species, and hence have a particularly damaging effect. Eradication of rats etc. on oceanic islands is a particularly successful example of (2), although for many species full eradication is unlikely to be successful. Since the same control would have to be carried out anyway (e.g. rhododendron control in non-wild woodlands), I hardly see this as an argument against rewilding.
A Britain without active conservation management would be covered by woodland, and ultimately this would lead to a decrease in biodiversity - we would gain lots more Nuthatches and Great Spotted Woodpeckers but no additional species, yet would lose numerous upland waders, Stone Curlew etc as their habittas became wooded. These have not 'always been here' - they probably colonised after neolithic man did some habitat creation and maintenance, like an unwitting stone age RSPB.
This is ridiculous extrapolation. No-one is suggesting that the whole of Britain be rewilded, and with a native woodland cover of just a few percent (and virtually zero that could really be called 'wild', this just isn't a realistic prospect. Tree cover in the uplands (i.e. the most likely areas for rewilding) could be increased massively while still leaving vast acreages of open habitats for waders and the like. Even within a rewilded forest-type environment, it's very likely that a lot of the moorland species would persist either in the understorey, in gaps, on cliffs and in wetter areas. Besides which, upland (rain)forest is inherently a much more species-rich habitat than the current vegetation it would replace.
Also, there is no reason why current conservation practices on small local nature reserves have to be abandoned - rewilding is obviously going to work better over large areas, so there's little to be gained by allowing isolated reserves to succeed to isolated patches of woodland.
By 'natural state' all you mean is unmanaged. There can never be 'natural' because we will never have the full range of species and processes going on. It will be a 'new natural' created by us. I see no logical difference at all between this kind of half-cocked 'natural' and a brand new RSPB wetland reserve. They are all gardening, and 'rewilders' are kidding nthemselves if they think theirs is a somehow more 'pure' and 'natural' approach. It's fantasy - you'll never have the Aurochs back and you'll always need to manage whatever hodge-podge of species you do have.
Actually no, unless you only use natural in a binary fashion (all/nothing). It is more useful to see degrees of naturalness, from a highly modified habitat (e.g. a wheat field) through e.g. a managed woodland, to a completely natural forest. See, for example, '
Natural Woodland' by Peterken, pp 14-15. The last is somewhat theoretical, in that nowhere on earth is now unaffected by humans, but is still a useful concept not only in rewilding but also in active management as something to aim for.