Changing your mind *after* the site has been spidered, ranked, and linked-to is worse. Then you have to deal with redirecting all the previously-indexed URLs, and hope that most of their link-juice gets passed from old to new.
I have a passing familiarity with mod_rewrite
[webmasterworld.com], so I'd like to point out that you'll be more comfortable with mod_rewrite if your 'mental picture' of its function is technically-correct: Mod_rewrite does not take "search-engine-unfriendly" URLs and "turn them into" anything. It is used (in this context) to associate a "friendly" URL with a script filepath, so that when a request for a friendly URL is received by the server, it invokes that script (and can pass parameters to it) in order to generate the correct content for that requested URL.
In this case, it is functioning as a requested-URL-to-server-filepath translator -- one of its three major functions.
The "changing of the URL" from "unfriendly" to "friendly" occurs when you edit the on-page link on your Web page or modify the script or database that is used generate that on-page link. URLs are defined on the Web by links on pages, while filepaths are defined inside servers. To illustrate this, put up a link to a non-existent page or even a non-existent domain. That URL now exists and will get spidered, even though it does not resolve to any content on any server... It's a perfectly-valid URL, but it just doesn't lead anywhere.
We frequently see confusion in this regard, and understanding the true process really does help.