The fundamental legal issue we need to deal with is the ability to
absolutely identify the originator of the mail. Is that precluded by any
existing privacy laws? If not, identity would provide the means to
pursue financial recourse for wasted time and resources. If so, we have
a non-technical issue that may prevent any solution.
Too bad the bad ideas get much more air time than the good ones. Yesterday some really good points were brought up, today we're mostly rehashing the bad stuff.
About the law: current laws are unable to keep spam in check. Is this a problem with the law? I don't think so. A good percentage of all spam (but certainly not all of it) breaks existing laws. It seems unlikely that additional laws will make people who already operate outside the law change their behavior.
Another preposterous idea: charging money for sending email. Economics dictates that this will increase the overall cost of emailing by many orders of magnitude and there is no reasonable upgrade path from the current situation to the new one.
Both points are moot anyway as they fall outside the scope of the IETF's activities. The real question is whether the current protocols exhibit flaws that make the spam problem worse than it would be without those flaws; and whether improved protocols can be implemented and deployed at reasonable levels of effectiveness and efficiency.
It seems the answer to this was "no" five or six years ago. In the mean time, many things have changed. We now have more advanced techniques and more processing power at our disposal. Also, spamming in general has become much worse and many more children are online now, who are subjected to spam that isn't always "child friendly" to say the least. Maybe the answer is still "no" but the time is right to at least revisit the question.