I now have time to comment on this thread.
It goes back to points that are difficult, that I have often alluded to over the years. This does relate to the fact that bench testing of power amps with resistive loads is of limited use and hides a lot of problems.
Amp and speaker manufacturers would rather not talk about, and most others, for that matter as it is complex. Viewing a speakers impedance curve alone is also of limited use. You have to consider the relationship of current and voltage as well as impedance. This is the thorny issue of power factor. This gets us to the fact that the speaker amp interface not only makes a difference as to how amps sound, but more importantly whether a given speaker is likely to blow up an amp.
The OPs question is a classic case in point. The Magnepan speakers present an almost constant resistive load with phase angles close to zero throughout the range. So there will not be retained energy in the output devices. So even though the impedance may be 3 ohms, I would say those speakers are very unlikely to blow up his amp. I would not worry about the possibility of them doing so.
Since the inception of getting to know Peter Walker of Quad and tinkering with the early 405s, I have had a huge interest in these two issues. First is designing speakers less likely to blow amps up and secondly making amps more stable under a variety of loads. This was also a great interest of Peter Cooke's, founder of KEF. He carefully designed his crossovers to avoid amp busting loads. This is also a major reason that I think that amps actually belong in speakers. I firmly believe that speaker and amp should be designed as an integral unit.
I referenced this article previously. It is an important if difficult to understand issue. All of you would do well to try and understand it, and what it that that makes different speaker designs, and different speaker combinations potentially lethal to the amp.
So if your amp or receiver blows up more likely than not the responsibility lies with both amp and speaker, whatever the speaker manufacturer says the "nominal impedance" is. I don't think there was ever a more useless and misleading spec. than a quoted speaker impedance number.
I have not been very active here of late, as I'm busy working at our new home built. I will try and post an update on the projects there this weekend if I can.