Talk about creating a Province of Toronto has surged then and again, usually prompted by complaints that the Ontario provincial government is neglecting Torontonian interests, in infrastructure and government service investment, say, in favour of thsoe of a wider province. Others have proposed that Montréal be separated from Québec in the event of Québec secession. Talk of the city-region, a region centered upon a city characterized by a sort of economic and demographic unity, as the defining entity of the 21st century has been current for a while. Kenichi Ohmae's The End of the Nation State imagines the deconstruction of nation-states into much smaller subnational units, each having their own policies in order to maximize growth. Jane Jacobs, famed Toronto urbanist, went so far as to suggest that each unit could have its own currency, the better to exploit its particular niche.
Andrew Sancton's The Limits of Boundaries: Why city-regions cannot be self-governing shoots these ideas down simply be pointing out that the boundaries of these regions are far too narrow. He examines other city-regions and finds them lacking: the failure of the 1996 referendum on uniting Berlin with the Land of Brandenburg that surrounds it has forced the two Länder to establish unwieldy common planning boards, while the huge fuss over language rights for Francophones in the Flemish districts around Brussels and the question of these territories' ultimate fate has risked shattering the Belgian state. Sancton approves of the Community of Madrid, but notes that the Community's frontiers were specifically designed to include Madrid and its hinterland during the post-Franco democratic transition. Sancton also raises the very important point that the frontiers of city-regions move outwards as technology advances and transport becomes easier. At one point, Hamilton was an entity separate from Toronto; soon, London may be included. Ironically, enfranchising city-regions as levels of government would stifle the dynamism that makes them so productive. The traditional levels of government, he concludes, are large enough and stable enough to accomodate cities' needs through their economies of scale, perhaps with a bit of tinkering necessary but nothing that can't be maanged..
(And yes, I know that The Limits of Boundaries is a book of obvious relevance to--say--talk of the partition of California into several units. Guess why I picked it?)