Nate Atkinson and Ezra Friedman have posted this paper on SSRN. Here’s the abstract:

Should states retain plurality rule or adopt a different voting method? We study the class of sequential plurality procedures, covering all widely-used single-winner election methods and proposals (including plurality, top-two runoffs, and instant runoff voting). We compare methods by whether a majority prefers the outcome of one method to that of another across all possible preference profiles. We show that a top-two runoff is the only procedure that robustly dominates plurality rule. All other methods perform better in some cases and worse in others. We further characterize the full set of dominance relations among sequential plurality procedures.

Although much of the paper is technical, its main conclusion should be of interest to election reformers. Note: the paper is an examination of voting rules that prioritize first-choice preferences and does not include within its analysis Condorcet-based voting rules.