Yes, it's an interesting article. I do think you need to ask yourself exactly what problem you're trying to solve before just jumping in, though. One thing that can be very useful is an antenna/preselector that can get rid of "boomers" before they get to the amplifier. For example, a small tuned loop can be oriented to null an on-frequency interfering signal, and will have enough selectivity (assuming it's tuned) that off-frequency boomers will be much less problem. That doesn't work if you _must_ listen to everything all the time, though even then notching out one or two large signals can be useful. Because of the very high atmospheric noise on MW and even up into HF, there is seldom a need for a very good noise figure; the distortion performance takes precidence because of all the large signals around.
With a good antenna, there may be no need for an amplifier before the mixer at all, and it's possible to build mixers with input IP3s up around +50dBm -- there are people that read this group that know how to do that. Once past the mixer, you can filter, and if your _filter_ had low enough distortion and selects the signal you want while attenuating others, the rest of the amplifier chain doesn't have to be _quite_ as good.
I have an abiding interest in amplifiers with good distortion performance over just the band mentioned, though, and I can tell you that over most of it, you can do quite well with modern op amps. One example (but not the only one) is the circuit on the front page of the OPA847 data sheet. The OPA847 also has very respectably low noise. There are a few non-operational-amplifiers that do quite well, too. The IC amps generally have more noise, I think, than what's mentioned in the article (I haven't read it all -- yet), but they should be adequate for at least MF and lower HF work.
Just this morning I came across a PDF talking about using simulation software to optimize an amplifier for good distortion performance--it was an ap note that touted the (somewhat expensive for ham use) simulation software from that company. I didn't bother downloading it because it was out of both my price range (the software) and my frequency range (the ap note).
I'm curious about comments from BCB DXers--from what I've seen, it's hard to find signals you could listen to that are all that small, because there are so many stations on the air that every "channel" has something at least moderately large in it. For example, I see local stations here at about -30dBm on the antenna I'm using, and I can see signals every 10kHz across the whole band that are seldom lower than
-100dBm--just one or two that are lower than that and not by much. In that situation, a +40dBm TOI should be adequate: distortion from the
-30dBm signals would be 140dB below them, or 50dB below the smallest signals I might want to listen to. But maybe if I spent more time looking at other times of the day, I'd get different results...and certainly having a neighbor running a kW on the same HF band I'm trying to listen to would require much better performance.
Cheers, Tom