The upshot is that for a regional party to get elected, they have to build a seven to eight figure campaign war chest per seat, because the second it looks like they have a good chance of winning, the national parties will pull a Brinks truck full of nationally-sourced money into their own candidate's campaign headquarters.
And while it's certainly true that money alone doesn't guarantee results (see: Linda McMahon, etc.), if you're running a campaign on $500,000 and your opponent is spending $50,000,000, you're in serious trouble. And it's difficult for a regional party that can't even guarantee the ability to filibuster to raise that kind of cash.
What I think we're more likely to see is what we've been seeing: Not separate parties, but independents with their own personal wealth who are willing to spend it to get themselves elected, and who have enough to make it a fair fight when their money is undiluted by multiple races. The problem is that then the candidates are self-selected and there is no guarantee that they'll agree with each other much less any of the electorate.
(I also think you're overestimating the economies of scale in party fundraising: look how much now happens outside the parties themselves -- independent expenditures -- and directly by specific candidates. It's only at the margins in winnable races that national party funds come in, and even that now has to be weighed against the negatives the national brand can bring. The new pools of money -- from both mass crowdsourcing and wealthy crusaders -- care more about winning than 20th-century party labels... and in the national legislatures independents/other-parties can and will caucus with other groups for the purpose of parliamentary procedures.)
Now, we have national cable news and 365-day-a-year recreational poli-tainment. We have outrage/indignation/us-vs-them magnifiers like Twitter and social news. More issues are discussed – and political affiliations chosen – nationally.
At the level of the casual observer/participant (and the 'marginal voter), I think that's led to unhelpful dimensionality reduction, into coarser and more childlike political categories. And that's especially dragged down the Republican national brand, because for several reasons (including less sympathy in the media) their most-embarrassing regional characters do more damage elsewhere than those of the Democrats.
So the devolution to more than 2 brands I mentioned wouldn't mean a real enduring 'third party' – I agree, the US system is rigged against that. Rather, it'd be getting away from the national 'Democrat' and 'Republican' names when they're unhelpful oversimplifications, perhaps via differently-named, and more doctrinally varied, regional affiliates... a little like the 20th-Century General Motors model of many brands.