Should Speaker Manufactures provide a Frequency Response Graph?

 I agree with erik_squires.  A frequency response chart made in an anechoic room has little to do with how that speaker will sound in your room with your amplifier.  Besides, the impedance of the speaker is changing with frequency which will affect how your amplifier/speaker cables sound coupled to those speakers.

Back in the early 1990's I built my first dedicated listening room and I added sound dampening to make my room have a flat frequency response.  The room was very close to anechoic.  The music sounded dead and lifeless.  I had to remove much of the dampening material.  

That's my point, I guess.  If you want flat frequency response then you will need a hemianechoic room like the room where they test speakers.  It will not sound so good.  Think about live music.  No band ever plays in an anechoic room.  Life is messy and so is music.

Most garbage speakers I have seen (and heard) had 3 things in common:  Cheap connectors, a little tab to pull off the velcro attached front grille, and a frequency response graph showing flat response from 30Hz to 22kHz.

Some of the responses here display the anti-science attitude which I find very frustrating in this forum. The vast majority of speaker manufacturers will measure the frequency response of their products, as well as a host of other parameters. An estimated in-room response (based on a large number of on and off axis frequency response measurements), such as that measured by Klippel, actually will do a pretty good job of indicating how the speaker will sound in an "average" room. It doesn't tell you everything, of course...distortion, directivity, impedance,  sensitivity...all are important. I wouldn't buy any speakers based on data without listening, but I would certainly rule some out based on data.

"As engineer for 20+ years I can assure you 99% of the high tech things you use (or even keep you alive, I work in the biopharmaceutical industy) (sic) were build or designed using models."
I do not doubt this, however I do not need to know and am not interested in measurements; I am interested in how they sound to me and whether I like that sound.

"Some of the responses here display the anti-science attitude which I find very frustrating in this forum." - see my comment above.

A major consideration where run of the mill measurements completely miss the mark is the materials used in the drivers themselves. For example, Kevlar, polyprop, etc used in drivers that measure perfectly...you might as well throw a blanket on your speaker before you started listening.

This is one of the reasons i don’t like the low effort/resources brands who buy mass market drivers from someone else and put it in a box. It takes some guy who has put serious thought into the materials used in drivers and developed his own from scratch.

We could go on about crossovers, etc... barrel bottom quality components in crossovers can still measure perfectly. With sim software, you can very quickly design great crossovers, But, as you move up in quality of components used, the sound totally changes, i.e. the measurements say little about "quality" (whatever that word means to you).

Last, but not the least, humans perceive way too much in transients (spatial nuance, etc). The configuration (Drivers+crossovers+box) that gets botched w.r.t the latter, no matter how great it measures with traditional measurements will sound like a ..."meh".

The above mentioned perhaps falls into the "esoterica" (dismissed) category for the ASR type of guy, another reason i despise that loathsome forum.