Q. I don't think that Range Voting in primaries and caucuses like Iowa will be an advantage to the parties. The candidate who will win the general election must be able to win a plurality vote.
A. Partially correct but unfortunately not fully correct thinking.
In Iowa caucuses, the candidate who wins is not necessarily a good one. Here by "good" we mean either in terms of quality, or in terms of ability to win a later plurality election versus the other team. The reason for that failure to generate a good candidate could be, say, essentially random effects due to strategic voting (which is a much greater distortionary effect under plurality than range voting.)
Example: Kerry won the Iowa Democratic caucuses in 2004. Was he really the best Dem from the standpoint of either electablity versus Bush, or in terms of quality?
Example: Bush won the Iowa GOP caucuses in 2000. (2nd was Forbes, 3rd was Keyes.) Bush then "lost" versus Gore (popular vote nationwide, anyhow). Was Bush really the best Republican from the standpoint of either electability versus Gore, or in terms of quality?
McCain was running too, and nationwide polls indicated McCain had more support than either Bush, or Gore. There are also reasons to suspect McCain would have been better quality-wise. For example McCain was a Vietnam Vet, has a record of urging fiscal responsibility, and has a record of being able to build bipartisan compromises. All those would have come in handy as president. So all the evidence suggests McCain would have been both more electable, and better quality.
So why did Iowans choose Bush #1, Forbes #2, Keyes #3? Well, McCain did not campaign in Iowa. Therefore, Iowans figured (strategically) he had no chance so there was no point to wasting their vote on McCain. Strategically, you "have to" vote for the top 2: Bush & Forbes. It was all about strategy, not about quality. It should be obvious that McCain was better qualified than Forbes and especially Keyes, both of whom have never held elected office. With range voting, while McCain might not have won Iowa, for sure he would have done a lot better. Because voters could afford to be honest about him in their vote-rating instead of strategically refusing to vote McCain. The result could have been better for Iowa, and the GOP, and the USA, and the world.
So the reasoning that primaries have to be plurality-based, is wrong. It causes strategic vote distortions that hurt the party that employs plurality primaries and lessens their chances to later win the general election and hurts the whole country too.
If the early primaries are Range Voting, then the top 2 from range effectively then are put into a head-to-head runoff versus each other by the remaining plurality part of the primary. (And note, that "runoff" gives them plenty of experience in a 2-choice plurality election – don't worry about that.) This results in far superior average winner quality – the best get elected, not those who due to random early strategic perceptions, get an early lead.