It's beginning to look like: (Democratic) interventions in ex-Yugoslavia and now Libya were well thought out, targetted and focussed, limited, and successful. (Republican) interventions in Afghanistan and Iraq were the opposite in each case, and unsuccessful. Even if an Iraqi democracy survives, it is apparently pro-Iran, pro-Assad in Syria, tied to Islamic fundamentalism (and potentially more supportive of the kind of international terrorism the U.S. is concerned about that Saddam was); it has come about by means of ethnic cleansing and the rise of sectarianism. Baghdad is a less cosmopolitan place, with a worse university, and fewer choices open to women, than was the case before.
Of course there is a huge contribution of chance in all these events. NATO had not claimed to be saving the whole world in ex-Yugoslavia, and the scrutiny of the world did not demand miracles. Ex-Yugoslavs who were given a chance to build democracy, without being pushed, did so; it must always be partly a matter of luck whether such people are available or not. But the Bush administration was extreme in its hubris: trying to create a popular uprising, instead of carefully supporting one that is underway; trusting Ahmed Chalabi, instead of ensuring there is some kind of cross-section of leadership actually available.
I'm inspired by an old article by Nathan Tarcov. How to intervene for democracy in a foreign country (or, more subtly, to help "a people" take control of the country)? Wait until a popular uprising, at least plausibly speaking for the people, is well underway; intervene strategically, only as much as necessary, preferably as part of a coalition of some kind (the role of the Arab League was crucial in Libya).