James Mann, whose book The Rise of the Vulcans: The History of Bush's War Cabinet (2004) is a very useful account of the backgrounds of a number of key players in the first years of this Bush administration, is writing about The curious disconnect in US foreign policyFinancial Times 04/16/06:
In its first national security strategy in mid-2002, drafted under the direction of Ms Rice, then national security adviser, and her deputy, Stephen Hadley, now Mr Bush's national security adviser, the administration summarised its policies with a key phrase: the US would seek what was called a "balance of power that favours human freedom". Those words represented a classic compromise - or call it a truce - between the party's two warring wings: "balance of power" for the realists, "human freedom" for the neo-cons.
The 2002 document contained some far-reaching ideas about dealing with terrorism - including, most prominently, the call for pre-emptive military attack. Yet, outside the Middle East, the administration seemed to view the world in conventional Kissingerian terms: stability, national interests, balance of power. The 2002 strategy singled out China, Russia and India as three centres of global power.
The new national security strategy is strikingly different. The phrase "balance of power that favours human freedom" has been dropped. There is quite a bit about freedom and spreading democracy, but not about the realist concept of a balance of power. Gone is the section that four years ago grouped China, Russia and India as great powers. They are treated in the 2006 document as three countries among many. For the first time, the US seems to be saying its power is so great that there can be - and need be - no balance or stability.
Mann considers that shift in official strategy "a conceptual change of breathtaking magnitude". I'm not so sure that the actual policies represent such a shift. It's probably more that the official strategy language is catching up to the reality of how committed this administration is to unilateralism and preventive war.
One reason for the shift in official language is that spreading democracy has become the rationale for the disaster known as the Iraq War, though that was not one of the two goals of the war as stated in the October 2002 Congressional war resolution.
Mann seems to have made the assumption that Condi-Condi's appointment as Secretary of State and some other personnel shifts, like Paul Wolfowitz' exit from the Pentagon, meant a decline in the influence of the "neoconservatives" and an increase for the "realists". Now he says that Condi-Condi and other supposed "realists" are promoting "now regularly espouse the ideas first put forward by the neo-cons".
I'm not at all convinced from what I've read of Condi-Condi's tenure as National Security Adviser that she was ever such a committed "realist". It sounds like she was more Bush's enabler than anything else.
But part of the problem may also be that categories like "realist" and "neoconservative" may encourage a notion of a left/right split, or a pro-negotiation/anti-negotiation split, or something of that sort.
Ivo Daalder last year suggested that "nationalism" is more of the unifying concept for Bush's foreign policy (Nationalists Once Again Triumphant Center for American Progress 03/15/05):
In the most recent battles between the Bush administration's realists and neoconservatives  over Iran, the U.N. ambassadorship, and jurisdiction of the World Court - the nationalists have once again emerged victorious. This should not come as a surprise. For realists and neoconservatives have always been a minority in an administration in which nationalists, from the president on down, hold all the major positions of power.
We should also remember that "realists" can range from genuine pragmatists to Machiavellian manipulators like Henry Kissinger to cynical operators like Dick Cheney. And no matter how pragmatic one's outlook on policy, if they are groosly overestimating American military capabilities, as the Bush administration consistently has done in the Iraq War, decisions are still likely to made that will produce results very different than those desired.