It's not easy for an alternative party like the LP to do large-scale empirical research on possible political strategies. Exit polls ignore us, and government officials don't like to re-run election-day procedures to answer empirical questions (unless they're Democrat appointees on the Florida Supreme Court). So if, for example, you wanted to to use election results to measure the anti-war issue as a way to grow the LP, you would need back-to-back elections in which 1) the first election established a relatively large baseline for the available third-party vote in a war-free context, and 2) the second election was a wartime election between two major-party candidates who both favored continuing the war and who faced the same alternative candidate who had established most of the previous baseline.
Amazingly, this is precisely the scenario that happened in 2000 and 2004. Ralph Nader received 2.9M votes in 2000 from voters who were presumably anti-war and clearly willing to vote third-party. Another 450K votes went to Pat Buchanan, the darling of anti-war ex-LP paleolibertarians like Murray Rothbard and antiwar.com's Justin Raimondo, Eric Garris and Colin Hunter (who all had been leaders with Rothbard of the LP's Radical Caucus in the early 1980s). Those 3.35M votes should have been available to the anti-war candidates in 2004, because both major-party candidates favored continuing the war that had started between the elections. Together Nader and Green rival Cobb reclaimed at most 580K of those 3.35M, but Badnarik improved on the 2000 LP presidential vote by only 13K! Even the tiny Constitution Party was able to add 46K to its total, a nearly 50% increase. Thus when about 2.6M anti-war third-party voters rejected Nader/Reform/Constitution to choose among 1) the pro-war major parties, 2) the anti-war LP, and 3) NOTA, the LP attracted only about 1% of them-- whereas 100% of them would still have been only 2% of all voters! It's hard to imagine better empirical evidence that emphasizing our anti-war position is not the best way grow the LP.
Understanding the origin and psychology of that position helps explain why the LP emphasizes it over others -- like privatizing education and healthcare -- that have no Green/Dem/GOP competition and that will still correspond to growing national crises long after lame-duck Bush and his entanglement in an Iraqi civil war are a distant memory. A large component of the explanation is the sincere belief that governments throughout the world, and even the American government throughout its history, have demonstrated insufficient ability to reliably choose military interventions that increase liberty instead of decrease it. Another significant component is that the Vietnam War and the Draft were monumental issues to the generation that founded the LP, and the institutional anti-war memory remained in the LP even as important radical anti-war leaders left it. (Some, like Rothbard lieutenants Bill Evers and Emil Franzi, later ended up supporting the Iraq war in their post-anti-war incarnations.)
But there is a fundamental ideological reason why opposition to war is considered by LP radicals to be the most important issue. The military defense of liberty is the textbook example of what in economics is called a "public good" -- a good that markets will underproduce due to the Free Rider Problem and that thus needs tax financing. Anarcholibertarian dogma denies this textbook market failure, and so zero-coercion absolutists have a deep need to deny that any net good could ever be done by a tax-financed military.
This could explain much of why anarcholibertarians are so peculiarly revisionist about e.g. the Civil War and World War II. It would undermine my hypothesis if we could find significant examples of 1) revisionists about these two wars who don't oppose all tax-financing of defense, or 2) opponents of all tax-financing of defense who admit that a tax-financed military could do or has done some net good in a war. The only possible example I know of is Ron Paul, whose borderline exception proves the rule. Paul defends the use of some coercive taxes (excise, not income) to finance the military -- although you might not know if from his recent appearance on the Daily Show, where even a TV comedian was able to instantly identify the national defense exception to Paul's blanket call for privatization, and force the former LP presidential candidate Paul into a muddled backpedal. Paul is indeed something of a war revisionist, but he voted to approve the use of the tax-financed military against those nations that aided or harbored the planners of the 9/11 attack -- and is criticized for doing so by zero-coercion anti-state absolutists.
As Lawrence Samuels wrote in CF: "To accept the legitimacy of the state is to embrace the necessity for war." On antiwar.com, Samuels opined that "technically, any taxpayer paying federal taxes can be considered an accessory to murder if they support involuntary taxation." Samuels resoundingly defeated the arguments of strawmen who claim "any resulting collateral damage is considered acceptable because the war is being waged for a good cause". In response to Samuels' quote of the Bourne aphorism that "war is the health of the state", Saddam Hussein's neck might reply that aversion to war is the health of the tyrant. Anti-war absolutists argue that wars have been a ratchet mechanism for the size of government, but a graph of the ratio of government spending to GDP shows otherwise. It shows only two short-term doublings: the peacetime doubling in the early Great Depression was sustained, but the WW II spike was not. Subsequent wars disappear in the noise of the growth of the nanny state.
Samuels said that "Iraq has never had a democratic government in its 7,000-year history, or possessed any grassroots movements advancing the concept of liberty." But America had in Kurdish Iraq a ten-year existence proof that the U.S. military could replace Saddam's tyranny with an commendable level of stability, prosperity, and self-determination. America achieved its objectives of 1) eliminating any WMD capability or international terrorist infrastructure, and 2) deposing Saddam's regime in favor of a federal democratic constitutional framework designed to protect minorities and fundamental human rights. By early 2004, polls showed that a majority of Iraqis believed "things were better now than they were before the war" and that "Saddam Hussein's ouster made it worth any hardships." What these Iraqis (and Cassandras in America) failed to predict was that sectarian strife was to develop into a bloody Sunni-Shia civil war and negate much of the value of achieving objective #2.
Reasonable liberty-lovers can disagree over whether a conjunction of self-defense and humanitarian arguments could justify the U.S. overthrow of Saddam. The crucial question is whether the duty of a liberty-loving polity to defend human liberty vanishes completely at lines drawn on maps by statists. Some very principled libertarians say it does, but it's hard to see how that position is liberty-maximizing.