I know nothing about FTTC, have never even seen it, but I have not heard about dynamic spectral masks; I could read through the entire of G.993.2 (VDSL2) but I don’t have the energy, I’m afraid.
So you’re getting more hissy noise bursts then probably, pushing the (downstream?) SNRM up that far. The QLN improvement could indeed be a crosstalker going away, as you say, or someone has turned off some kind of electrical equipment that is a noise source.
It has got noticeably colder here, don’t know what it’s like where you are. The lower temperatures have reduced the attenuation figures on my lines due to lowered electrical resistance, and this has meant faster DSL, up by a total of about 0.6 Mbps to 10.2 Mbps combined total TCP speed tester downstream result / 1.0 - 1.4 Mbps upstream (depending on which speed tester you look at). So I’m wondering if reduced temperatures are having a beneficial effect for you too.