In general, if we expect things to work we won't get into much detail intentionally. If we are transparent about things (explaining just in case we have created this expectation), it's usually in the context of explaining what we are doing about an issue for the sake of productive communication with the community or to give some technical context on things so that the feedback that we get on a particular area is more focused. Ideally things just work so that people don't even need to discuss it. There're reasons for this discretion that involve often-unfounded anticheat fears, competitive-edge protection, not giving fodder to patent trolls, time constraints, embarrassment because an issue was too dumb, a desire for conciseness to not waste people's time or a general sense of modesty. Also netcode may be the kind of thing where you don't want to explain too much or people start reporting ghosts because you colour their experience introducing concepts for which the scope can't be completely understood.

Having said that, if you must know, the context of this iteration on netcode was this thread:

https://www.reddit.com/r/Diabotical/comments/j248ug/ive_realized_i_have_no_desire_to_play_this_game/

We had been seeing these things here and there and although I never discard a blunder in the implementation my main suspicion was packet loss episodes since we never managed to reproduce these while testing, and I was waiting to have a lagometer so that these reports could include that info. (Sorry, that is coming at some point, but we've had other priorities). Nonetheless, since I got the feeling this particular user wouldn't be that frustrated if there wasn't a bug involved, we went ahead assuming it wasn't lag and did more testing and developed some internal analysis tools and found one small area for improvement. This also allowed us to implement an improvement that we had in the back-burner that for some reason was failing some tests and it was due to this same issue. Those things combined should largely prevent situations like the one in that video where rockets are going through people, unless there's a lag hiccup involved.

The issue we found was handicapping the accuracy of that initial segment of the projectile journey in very close player vs player situation, as some people suspected in that thread, though they are attributing those to inter-player collision issues. Inter-player collision is wonky but it should not create the kind of issues that people tend to classify as netcode issues. We'll work on that too.

The general accuracy fix should result in an increased ability to hit at any distance (this may casually manifest as hitting more air rockets since that's the thing people are most likely to remember).

We will probably always keep improving netcode for as long as the game is maintained. I do expect, though, that things should be at a pretty playable state after this patch, registration-wise.