Adam B. at Daily Kos has a very good post on what accepting public financing means for John Edwards.
I am going to try to summarize and correct where I think he is wrong. He is a lawyer and I am not, but I still think I'm correct.
1. The overall expenditure cap for Edwards will theoretically be a little over $42 million. He can also spend up to 20% on top of that ($8.5 million or thereabouts) for fundraising expenses. The definition of fundraising expenses is pretty broad, so he would certainly max that number out, giving him an effective cap of maybe $51 million. Any money he spends on legal stuff (compliance, etc.) doesn't count at all.
2. His Iowa maximum expenditure is theoretically going to be around $1.5 million. However, this only counts media expenditures, mailings, overhead (rent, furnishings, utilities, phones) phone programs (telemarketing and/or robocalls), and in-state polling. For media it only counts the percentage of the buy that will reach Iowans, for overhead it only counts 90% of expenditures, and of that remaining amount only 50% of it really counts.
To put that in equation form, it would look something like this:
(Iowa% * Media) + (90% * Overhead) + Paid Phones + Mailings + Polling <= $3 million.
That's a very low number, but then again Edwards can just ignore the rule and pay some fines later on (probably after the end of the general election, even). Therefore any concern about the limit is misplaced, and he should have plenty of money to implement his early-state strategy. Of course, early-state strategies rarely work.
So, while this isn't as bad as I first though, I'm still of the opinion that an Edwards nomination turns a Democratic advantage into a Republican advantage for the general election. I would say he's still more electable than Kucinich and Gravel, but has dropped far below every one of the other five major candidates.