Since no one else has responded, I will. I do in-room measurements often.
Figure 1: Frequency response:
Looks reasonable, no big peaks or dips to contend with are obvious at the 1/3 smoothing selected. However, you might want to use no smoothing in the low bass as to not lose the detail in that area.
Figure 2: Impulse Response:
Why is the first millisecond so hashy? It is far from ideal. I use room correction, but even my uncorrected IR looks similar to my "best example" - first two milliseconds:
The apparent levels of the later reflections may be accentuated by the lack of a strong peak in the initial response in the measurement.
Here is a longer IR from a couple of days later which shows the major reflections in my room using dipoles:
1 ms: probably reflections off the top of the couch where the microphone is
7 ms: the backwave off the wall behind the speakers
27 ms: double room-length bounce
If the initial peak on mine were more smeared out (like yours appears) the highest peak would be of lower amplitude and the relative amplitude of the reflections would be higher.
Figure 3: Waterfall Decay time
Your concern about the 20Hz area...
You might want to take a measurement without much input from the speakers, just the ambient noise levels. The lowest frequencies can have quite a high SPL, and while not being particularly audible, are quite measurable.
Here is a Waterfall for my room, with the excitation sweep at -60dB - you can see a little waterfall at, say, 4kHz, but the rest of the sweep would be below the level of ambient noise in the room (or measurement system) so it isn't even seen here:
The point being - it may not be so much a lack of decay (at 20hz), but just mostly ambient noise that you see in the low frequencies of your waterfall. I have a 18hz bump when the AC is on, and plenty of other LF content intruding from the outside world.
So, how much of that 20Hz in yours is always there, and how much of it is lack of decay from your test tone?
I'm not saying my room is noisy - all I hear is a little muted computer fan noise, but the measurement microphones are quite sensitive. Right now, 2am., 35.6dB A-weighted, 44.5dB C-weighted, 50.9dB Z or unweighted
Figure 4: The System
Very nice. I simply must do something better with mine... one of these days...
---
Sorry if these are not the responses you would have liked to have seen, but, at least it is a reply, based on my recent measuring escapades.