How thick should the front baffle of speakers be?

wheres the proof

In extensive notebooks and old hard drives.

I just contacted my friend (he's now a major audio/video integrator in L.A.). He reminded me, that not only were the improvements audible, but they were measurable. 

You do understand that there are many tests that can be run on speakers, drivers, cabinets, crossovers, etc, that have direct correlations with sound quality, right?

Pulse tests and waterfall plots were noticeably better in deader, less resonant cabinets.

I don't even understand why this has to be explained? Build a cabinet with less resonance, that is better damped, and the entire speaker system has less resonance, and is better damped. 

The fact that YOU do not understand the types of things that make speakers sound better, is not my concern. The fact that you don't even own a system, and are commenting the findings (subjective and objective) is quite laughable.

3/4" plywood didnt work just fine. Resonance is a bad thing. Think of the surface area of a large speaker cabinet much if not all of which is vibrating. 

   Hi, The speakers that I own (Sonist-"Concerto 4") have two inch thick poplar wood front baffles.    I talked with the designer (Randy Bankert) and he said he tried a bunch of different woods before settling with poplar as it was the best sounding to his ears.
   Instead of mounting the drivers on the outside ,he mounted them on the inside and beveled the wood to create a horn of sorts to help with dispersion. If you look at a picture of them you will see what I mean.
TISH

I've heard the Concerto 4 speakers at an audio show!

They are extremely good, for sure. When I heard them, knowing what I do about speaker cabinets, I made notice of their cabinet construction, and how that contributed to their sound quality. 

The rear mounting with beveled edges is a waveguide, which is a type of horn. 

More traditional horns, tend to be more concerned with gain as first priority, where waveguides tend to be concerned with controlling the pattern of dispersion as first priority.