it depends,every case is different.in my case i have a lot of FEC 80k a day when gaming,normally 40k,means that errors get corrected,that's what increase latency,
No - FECs are errors that never happened, and take no time whatsoever to fix. If you get zero, or 5 million, you will see no change in latency.
Interleaving is what adds latency - and that happens from the moment the line profile is set that way. Once the line is configured with interleaving, then even with zero FECs, you will have encountered the impact on latency.
looking at a table here Burst Protection adds huge latency.
The important point, however, is that the mere existence of "interleaving" does not imply there is extra "huge latency".
The table you quoted applies to ADSL, and the relationship between "latency" and "interleave depth" really depends on the total speed.
In VDSL2, the important parameter is the "delay" parameter that you highlighted in your first screenshot. There, your line was allowing 8ms of extra latency downstream and 3ms of extra latency upstream: a total of 11ms extra latency.
To achieve this, the modems agreed an interleaving depth of 837 downstream (and an interleaved block size of 128; this is an important number too), and an interleaving depth of 285 upstream (and an interleaved block size of 44).
However, go back and look at the same delay value when G.INP is running: a "delay" of 0. Zero.
Here, I have G.INP running, with INP=48, delay=0, interleaving depth=16, and interleaving block size=139. Your downstream settings (when in mode 2) allow for 8ms of latency, mine (in mode 3) allows for 0.16ms.
As you saw from the graph yesterday, my round-trip ping times (home - London) are currently around 13ms; the extra latency from G.INP is approximately 1% of that round trip time. I wouldn't classify that as "huge latency".
If you post the same screenshot of your line profile with G.INP running, I can calculate yours too.
