As a professional Acoustic consultant, I would give few suggestion about measurement tecnique (even if it is quite hard, with my poor English). First of all, the way you took measurements cannot give reliable results. In my opinion, the best signal to be used is pink noise. If you have a SLM with freq. analysis you can use a full band pink noise. Otherwise, there are CD with pink noise filtered in 1/3 of 1/1 octave bands. If you use a pure tone, the measure will be too much affected by the mic position (standing waves). If you repeat the measure 5 inches away, you could obtain totally different curve. In any case, the measure should be always taken averaging the value on a zone close to listening position, and not in a fixed point. If you have an integrating sound level meter, it will make the average by itself, otherwise you can average by yourself, but always using SLOW time response.
Regarding your curve, the response in low freq. is normal for a typical listening room. But, the high freq. response is usually much more flat, and your curve is probably affected my measurement errors.
As Abstract7 says, it is quite hard to get room correction with passive device at low freq. due to the long wavelength. Any absorbing device is effective if its thikness is at least 1/4 of the wavelength. At low freq. only resonant devices are effective, as Helmoltz resonators, or "vibrating board" resonator (I don't think this is the correct english transation). But, both of them, to have a good effect, requires a lot of surfaces and volume, (and have a very low W.A.F. :-).
For high freq. absorbtion, I suggest to avoid any exotic and expensive devices. Normal carpets and curtains are more than enough.
