Interesting - sounds like you are all basically saying that acoustical treatment is overrated, whereas others have said its vastly underrated.
I'm clueless, but interesting that there is such a divergence of opinion.
Let me put it this way - is acoustical treatment necessary for good sound in a fairly small room, or one of those "golden ear" things that certain people do in pursuit of absolute perfection?
Really, this depends on specific circumstances.
If you happen to have a room that is populated by furniture, etc. that provides a reasonable level of diffusion and/or absorption, then a speaker with even off axis response may sound good in this room. If however, you have a room that has virtually nothing in it other than a chair or couch and a table - and has serious flutter echos, etc.; then treatments would be needed to get that room to sound good even with the speaker that has a good even off axis response characteristic, at least for recordings that have reverberation/delay times lower than the delay / reverb times produced from the room itself. Recordings with delay/reverb times exceeding the room's, will tend to mask the room's own properties in this regard. But many recordings, especially studio recordings, are going to have relative short delay/reverb times much of the time. Also, if you place a speaker in a position that yields a short delay(less than about 4-5ms min.), high intensity reflection, you may need to use specific localized treatments to correct this non optimal placement effect. This would usually only be an issue placing a speaker directly next to side wall, or placing a speaker that has rear-radiation, very close to a rear wall.
In special cases, room treatment can be used to optimize a room for specific acoustic objectives that would otherwise not be possible; and only specific treatments may bring about these objectives - but this is really not relevant to the average situation - I just thought I would mention this as a disclaimer.
Additionally, LF response is usually problematic. Treatments designed specifically to target LF response problems can have a very favorable effect - unfortunately you need a relatively large number of them to have a substantial effect on LF response. You can also use a parametric EQ to help improve LF room characteristics; at least in specific seated position(s).
-Chris