Speaker impedance patterns - how to read them?

I listen to speakers or read reviews of what they sound like and ignore all but a quick look at a couple parameters. I was a former scientist and analytical person who has spent his career doing highly analytical work. I found most the analytical stuff just leads me away from sonic truth. Not wrong to be interested in it… but for me life is too short.

"...Not wrong to be interested in it… but for me life is too short.."

Exactly this. Impedance matching the speakers and amp is more about how they sound rather than having a perfect impedance match on paper.

The question posed by @tomic601 was the exact one Roger Modjeski asked when a potential Music Reference customer inquired about the suitability of one of Roger's amps for use with a given loudspeaker. He advised taking a SPL reading at the preferred listening level, a measurement of the amp's output voltage at that level, then extrapolating the required increased power requirements for every 3dB of added SPL.

I have been conversing with a fellow Eminent Technology LFT-8b owner, who has been marveling about how well his moderately-powered tube amp (I do not recall the amp, but it's under 100w/ch) drives the 83dB sensitivity LFT-8b. That's the same sensitivity as that of Maggies, which require and benefit from a high current amp. Sensitivity alone tells one only so much about a loudspeaker.

Maggies are a nominal 4 ohm load (dropping to 3 ohms at some frequencies), the LFT-8b 8 ohm. And if you bi-amp the ET (the speaker is fitted with dual binding posts, one for the m-p panels the other for the sealed box dynamic woofer), the magnetic-planar drivers present an 11 ohm load to the amp, great for tubes. By the way, the owner prefers the LFT-8b to the Maggie 3.7 he previously owned. The most under appreciated, under-acknowledged loudspeaker on the market. $2499/pr. VPI's Harry Weisfeld declared the LFT-8b to have the best midrange he has ever heard, regardless of price. Yet it continues to be ignored by most. Brooks Berdan chose Eminent Technology as his shops' magnetic-planar loudspeaker. 

So here you go:
Most conventional cone speakers have a rising impedance as frequency goes up. What most speaker designers should do that do not want to impedance compensate their designs is rate the impedance based upon the operating range of the individual drivers. Most companies do fudge numbers, if a speaker that they are currently calling an 8 ohm speaker showed that dipped into 3 or 4 ohms, many tube aficionado’s would not buy them. A realistic example would be: if you have a 2 way and the woofer is rated at 8 ohms but actually varies from 4 at dc to 30 ohms at 4000 hz. Realistically that woofer will range from say 6 ohms at its 3db down point of say 35 hz or so to a 15 ohms at a crossover point of 2k.... So in the woofers operating range, it varies from 6 to 15 ohms. it is called an 8 ohm woofer.
Rising impedance can easily be controlled. So in the example above, that 30 ohms could be controlled to stay within 6 to 8 ohms. My last design is a 15 inch 2 way, it is 95 db in sensitivity and varies from 5 to 6 ohms at all frequencies and yes, even though a solid state amplifier handles rising impedance easier, it is still better for an amplifier to see a very consistent load and if your tubed amplifier had 4 ohm taps, it would much prefer a impedance compensated speaker. Normally overall, phase angles are tube amp killers as much as just impedance, a tubed amp lover should look at phase angles as well as impedance curves.
There are the basics with a very basic example. There are times that other scenario’s could happen, but the above is the most common for conventional drivers. I hope this helps, Tim

Most conventional cone speakers have a rising impedance as frequency goes up.

I’m sorry that’s not really true. You are describing the impedance of a coil.

It is true that larger cone drivers have this, and that this effect needs to be dealt with but it is not true for a speaker as a whole. A quick survey of impedance curves at Stereophile would show this.

Here’s a random sample:

https://www.stereophile.com/content/stenheim-alumine-three-loudspeaker-measurements

It is also true that the most challenging areas for a conventional speaker tends to be in the bass, but the impedance curve of a raw woofer is not at all a good indicator of the impedance curve of the speaker as a whole.