how do they compare with your previous connection? (ADSL1 was it?)
your already getting more than the average 44db line at 8.2Mb..or are you forcing that by lowering the Noise Margin? if so then yes your FEC count will be higher, the lower the SNRM, the higher the error count. If I put my SNRM down to 3db the RSCorr (FEC) jumps up considerably, right now your FEC accounts for about 0.2% of all your traffic (i think) - I check the RS (all traffic) vs RSCorr (corrected) vs RSUnCorr (uncorrected, these cause CRC's) and roughly work out how much of my traffic is having to be corrected and compare - one thing I would like to know is just how much latency does correcting traffic add, i.e. how much slower are web pages and downloads if more of your traffic is having to be corrected vs un-corrected? - as far as I can tell it must be micro-seconds and why theres not much noticeable difference, only when the traffic cant be corrected and re-transmits occur.
what was your sync rate on ADSL1 (default noise margin)
and what is it on ADSL2+ (if your result above is with reduced noise margin)
some people on the border of where speeds may be better on ADSL2+ are usually better on ADSL1, your adding twice as many frequencies and these are "weaker" than the first 'half' that you use on ADSL1, and sharing the signal across them when noise is present just weakens the signal further, looking at your QLN it would seem that the noise starts around tone 120 and ends around 380, ADSL1 operates up to tone 256 and ADSL2+ extends it up to tone 512 - because the location of the noise in the tones it means that even though your doubling the use of tones in ADSL2+, your also 'roughly' doubling the 'introduced' noise (tones 257-380), but because these tones attenuate more and have a lower SNR than the lower tones of ADSL1 it means they are affected more by the noise resulting in less speed - also if you take a look at your QLN the noise seems slightly worse/stronger from tone 255-350 before lessening, these combined mean the ADSL2+ signal is taking much more of a hit on SNR and usable bits across those tones, this is why your not getting any/much increase from switching.
if you get 'some' increase and you dont notice any slowdown of traffic and web-pages still render at same speeds and dont 'hang' now and again, then it may be worth sticking with it and working on locating the source of the noise. One can get drawn into keep checking their xDSL stats and trying to keep sync high and errors low to the point it becomes obsessive, sometimes Im glad I cant check stats on my Sky Hub because I would be checking them all the time & trying different Noise Margins VS Error Counts etc lol... infact ive just purchased a Huawei HG612 so I can have a BCM6368 chipset with tweakable SNR and stats monitoring...lol
EDIT: I see you have an HG612...great, you can help me unlock it when it comes lol