How can you tell if a speaker has impedence dips?

Meech33,

Here is the link to the Stereophile review done in 1998:

http://www.stereophile.com/showarchives.cgi?628:3

I'm "sending" you directly to the measurement page showing the impedance & phase plot.

Lowest impedance is 4 Ohms @ approx. 200Hz (mid bass region). Music has a lot of presence in this region & it is a bit unfortuante that the impedance dips this low in that region! Your amps better have the power to compensate! This conclusion is, of course, based SOLELY by looking at measurement results. Many times results mean squat & the speaker is just fine w/ moderate power!

FWIW

Folks,
It's always a trade off. One amplifier would feel better when it has a normal load and another one would feel better working as a welding machine with literally short circuit(upto 1 Ohm dips) depending on power supply in general so the benefit might be better output characteristics for certain load impedance.
As to voice coils, the smaller the voice coil, smaller ressistance and more current demand to drive it.

My old 1980s vintage PS Model 2 amps sound WAY better bridged. It's not even close. I'm driving old Rega 2-way speakers with them.

It probably won't effect anything really. The Paradigm's are of such a load that an amp will find them easy to drive. The only way to know impedance for sure is a test whether a review or manufacturer. If the amp runs hot, simply parallel the speakers. This will keep the impedance up. A bridged amp will like this load better anyway. When you bridge, your amp will see an 8 ohm load as a 4 ohm, a 4 ohm as a 2 ohm and so forth.
I do agree with the above poster about deterioration in sound quality when bridged.

I will explain that in simple principal of work of dynamic speaker that consists of magnet, voice-coil and diffusor.
The voice coil just by itself is an inductor coiled arround the magnet and hooked up to the diffusor.
Any deviation of a voltage between the voice coil ends will result a mechanical motions along the magnetic core as per Faraday's Law and swing the diffusor that will transfer the mechanical motions onto the air.
Every voice coil has a number of coils that defines its inductance with Inductive Reactance and Resistance accordingly. A Capacitive Reactance between the coils might be neglected for impedance dips explaination. The larger the number of coils the larger the Inductive Reactance and the larger the Resistance as well.
A Reactance is Variable part if Impedance that depend on freequency. Resistance is Constant part of impedance if considered at the same temperature.
Inductive Reactance is directly proportional to the freequency i.e. whenever freequency rises the reactance grows hence we'll have an impedance rise; and the other way arround whenever the freequency goes down the inductive reactance goes down as well down to zero for DC. In case with DC we only have a resistance of voice coil that defines Impedance.
Resistance is a parameter that does NOT depend on freequency and in some cases may be defined as the LOWEST speaker impedance that can be mesurable with simple ohmmeter accross the speaker's binding posts(pair of binding posts per speaker mmust be connected with jumper or wires in that mesurment case) and so will define the lowest "dip".

I have read that some amps don't sound as good bridged, this could be another disadvantage if sound quality has to take a hit for more power.

I had some 20's a few years back.
They arent real hard to drive.
At the time I had a Dynaco ST-70 @35 wpc that gave me good but not great SPL's.
The bass on these little guys is very tight and go deep enough for most pop,light jazz or light rock recordings.

No dynamic cone speakers that I know of stay at a constant 8 ohms over the entire audible bandwidth. I'm sure the Paradigm's drop to 6 ohms or less at some point, but your best course of action would be to contact Paradigm. They have to have that data on file... You could also look for reviews. Sometimes they contain that information.