As far as my philosophy-of-ed naivete is concerned, perhaps the best example of a contemporary equivalent of whole-cloth curriculum theorizing in the vein of progressivism, Summerhill, etc., is Marion Brady's Seamless Curriculum. As someone who has taught at different levels but whose arguments don't explicitly come from any single intellectual root, Brady is iconoclastic and sometimes hard to read when he tries to squeeze his curriculum perspective into a shorter piece of writing. Disclosure: My difficulties may come less from Brady than from my own background, which is fairly far from curriculum theorizing. I'm aware of the conventional stuff (formal v. hidden v. taught v. tested v. ... curriculum) as well as the classic critical readings (e.g., Apple) and standard curriculum historiography (e.g., Kliebard).
My reading of Brady is that he's a nominalist: He places his view of the world in opposition to what he sees as a disciplinary slicing of the world, which looks remarkably like a modern version of realism (i.e., that there is a reality, and that the disciplines really do connect up with that underlying reality). He argues that reality is not that sliced up and, moreover, we can't teach children how to understand the world in that sliced-up way. Instead, he says, we should teach students in a way that matches up with the who-what-when-where-why questions (or the 5 Ws, for those familiar with journalism), which he says is more useful pedagogically. Maybe I'm mixing up my philosophies, but that seems very close to the nominalist position—that whatever reality may or may not exist, we tend to put categories and names on that reality in a very human way rather than a way that directly reflects reality.
But these days, there are many different varieties of nominalism. Just to name a few modern ones, there's pragmatism, where the categories are human and a good thing, too; there's deconstruction, where language serves as an inpenetrable barrier between categories and any reality; there's the radical science-studies field of Bruce Latour and others; there are the softier philosophers of science such as Ian Hacking, who simultaneously play with social constructionism and yet don't buy into Latour's and others' arguments; etc. And I suppose within each flavor, there are different attitudes one can take towards the nominalist position. Should we be regretful nominalists, who see dangers in however we slice up our description of reality? Should we be enthusiastic nominalists, who see the human classifications as potentially heuristic? I know I'm taking huge liberties with these concepts (the real philosophers who contribute to this blog will probably slap me with dinner-plate fish over this), but there's a point here... I'd place Brady as a regretful nominalist, though he may well disagree.
There are two weaknesses in Brady's argument. One is the inconsistency in his nominalist argument: Holism is as much of a construction as disciplinary silos. If the disciplines are an arbitrary division of reality, so are the 5 Ws. On what basis is the 5-Ws template a better one than academic disciplines? He asserts it's easier for children to use, but I'm not convinced. It may be easier for children of journalists to use, but young children at around four generally use only one of the W questions (why), and to my calloused parent ears it's not clear when that question functions as an interrogatory and when it's performative/interactive. Brady's ideas may also ignore the capacity for children to understand abstract ideas and categories (see Rick Garlikov's essays on teaching math to young children for another iconoclastic and very different approach). On the other hand, there's a long history of such underestimates, including Piaget.
The second weakness is the regretful approach to towards nominalism, the implication that just because academic disciplines may be artificial, that means that they're suspect for teaching children. My guess is that if you've gone to read Brady's online essay, my classification seemed right or wrong instantly. Why? Because the realism-nominalism duality has a clear meaning to those who've had some philosophy. (Those who have the professional expertise to slap me with dinner-plate fish over my errors will note that I'm ignoring conceptualism.) So having had some philosophy helps put Marion Brady's ideas into a larger context. But academic philosophy is a discipline, and if Brady is right, it's part of the artificial division of the world by disciplines. But it's helpful for explaining his rejection of philosophy as one of the disciplines.
I'm much more of an enthusiastic nominalist than Brady is. The disciplines are not perfect, but they provide useful perspectives, and to throw them out just because they're often used poorly or reified in schools today is tossing the baby out with the bathwater. That doesn't mean that cross-disciplinary themes/approaches can't be used—far from it, as Central Park East Secondary School's Habits of Mind (five organizing questions) comprised one example of a successful unifying approach. But that approach seems inherently interdisciplinary rather than being rooted in holism.