During development of an FFT based NVH course (wow...over 20 years ago), we did experiment with flywheels that were artificially made unbalanced. We learned a couple interesting things. It seemed that the flywheel must be balanced to itself because the crankshaft does not rotate at even speed, especially under power. The flywheel does its best to maintain its own inertia and the imbalance effect didn't do what we hoped.
In the process of taking stuff apart and putting it back together, we found (by accident) more vibration from a flywheel that is running out of plane, than one that is out of balance.
We were trying to create training conditions that the students could physically notice, while seeing an obvious result on the FFT. We actually got pretty far out there with added weight, but it just wasn't enough to make the engine shake like we thought it would.
We actually couldn't feel the out-of-plane vibration until we put a trans and clutch back into the vehicle. That spooked us because we thought we were done and were going to send the SUV back the fleet. It was one of those moments when you go, "uh oh....what the heck did we do?!"
Our problem was an un-noticed gouge on the bolt circle face of the crankshaft.
We never bothered to put those tables in the FFT program, and decided not to train techs to find primary crank vibration. You can fix a lot of cars that aren't broken if you have expensive tools to find things that aren't even there.
The one lesson, that was most useful, was that vibration you can feel MUST be working against something. Sometimes the transmission path is more of a problem than the vibration itself. I have seen engine vibration that was caused by poor bracing below the crankshaft centerline (back of the block) and the bottom of the bell housing. This is often a strange one, because the actual frequency is a rocking couple effect that does NOT match firing frequency events or crankshaft/flywheel rotation frequency (I have seen some that vibrate at 1/3rd of crankshaft RPM). Some engines need diagonal bracing of the lower circle of bell housing from a point forward of the rear block face (along sides of the block).
You can get a $15 app on an iPad or smartphone that will let you measure the actual frequency of your vibration. Find the actual vibration frequency at whatever RPM and then do the math to figure out what it really is. The sound frequency picked up by the mic, is the same as what your feel/hear. I don't think you need to invest in an expensive accelerometer and signal conditioning device.
One last note: We did teach the technicians a trick to change engine vibration amplitude by putting a paperclip in the OX1 terminal of the underhood check connector. When they held the paper clip with one hand, and put a finger of their other hand on the battery Positive post, the engine vibration diminished on the FFT. The test process was driving the fuel injection into a lean condition (using your body as a resistor from the 12 volt battery). The lean condition fires softer, which makes less vibration. This was really useful for identifying problems such as hardened radiator hoses (which shake the radiator in its mounts, which shakes the forward body baffle, which shakes the tail of the car, and makes the spare tire well boom).
One or two cylinders that fire harder (richer, higher compression, ignition issue...whatever) can have a pretty good shake going on.
Hope this helps somebody.