Is it good result with such low SNR margin (about 60-70 es per hour) on fast path?
I think only you will be able to say -- by
using the circuit and seeing if it allows an adequate service.
But this 415-473 gap is annoying.
Correct me if I am wrong but I thought your CP/ISP has already told you that the gap is to be expected. I know nothing about how your CP/ISP operates their service. I know nothing about the deployment of xDSL in Poland. I can understand that it is annoying to see the gap but if the service provider has said that it is an expected feature then there is nothing that can be done.
What do you think about negative SNR on some tones eg. -4.1875 on 455th tone?
Firstly, remember that you are looking at
ratio reported in decibels. A logarithmic scale. The fact that the value has a negative sign is just indicating that the ambient "noise" appears to be far greater than the perceived (but non-existent) signal. All you are seeing is the electronics of the modem/router attempting to measure and quantify "something" that is not actually present. We know that there is nothing there because your CP/ISP has previously said so. As far as I am concerned, there is no fault with the modem/router and there is no fault with your circuit . . .
Capture a full 24 hours of statistics and then post the graphs. Let us see how the circuit behaves in its current configuration.