Ok, I'm going to start trying to address the video, the specific times in the video might need to wait for the next post depending on how much time this takes to write. To anyone reading this forgive me if I explain something you already know I'm not trying to say you don't know it already I'm trying to distill this information so that anyone can understand. If I make an error in my explanation and you know better please correct me for the benefit of anyone who ever reads this.
I'll start by saying I'm not an authority on this subject but I like to think I have a decent general knowledge of this. I read a lot! and I have done some studying of newnet when I helped asosed for a short while with his FNN. You should be skeptical of anyone who claims to be an authority on these matters as there aren't too many left in the UT community who could actually be considered an authority. Those that are an authority will most likely be well known.
The server that this video/demo is from is running ZeroPingPlusv103 which as most know is a mutator that provides client side hit detection for hitscan weapons. That means, in this case, the client tells the server when it hit a target using primary fire of the combogib weapon. The combo ball damage/hits (secondary fire) and the hurt radius damage done by combo super explosions (primary fire exploding secondary balls) is still decided on the server with zeroping. Quick clarification on this special combo super explosion - the client tells the server it hit the combo ball with primary fire with zeroping active but the server decides where the combo ball is located when it exploded and how much damage is dealt to things (based on their location determined by the server) in the vicinity of this explosion.
Each client and the server are obviously running their own version of the UT game but the location of every player, combo ball, and other moving objects can differ on each game. The Internet is used to keep these positions in sync but because the internet protocols we all use are only a best effort service there can be delays or missed updates in these positions. When one of the running UT versions doesn't get an update or an update is delayed it just simulates where it thinks that object has moved to so that things seem to move smoothly. There is more to be said on this subject like the different delays based on the each players' distance from the server, and the servers ultimate authority on these positions etc. but I'll leave it at this more simple explanation.
All of this explanation was just to discuss what anomalies a player might see when playing on a server with this configuration. These anomalies are important to understand as they are often mistaken for possible cheats.
First, you might see an opponent and you run behind a wall to where he or she can no longer see you and then you die from that enemy's primary fire. This seems impossible but remember, with zeroping, the enemy's client tells the server he hit you before you ran behind the wall and the update to your client machine telling you that you died was just delayed somewhere along the way. It does not automatically mean someone is cheating when this happens.
You will often see what players commonly call unregs (unregistered hits). You shoot your primary fire and see that it directly hits someone maybe even multiple times but nothing happens. Well that shouldn't happen if your client decides the hits for hitscan weapons right? ZeroPing is well known for this issue it could be a bug or it could just be a delay or lost packet when your client sends its transmission to tell the server it hit something. This doesn't necessarily mean the player you shot is somehow hacking.
Also on this server, there is a mutator that plays hitsounds to the client when he hits something. Because hits are usually instant death in combogib we tend to associate these hitsounds as kill sounds but really they just indicate that you did some amount of damage to an opponent.
So another strange thing that you might have seen is you shoot one of the secondary killer balls at someone and you watch it land directly on them, hear the hitsound and then watch them run away. How did they survive? In this case the damage done by the combo ball is decided by the server based on the location that the ball hit your opponent. If you only just grazed him or her with the ball it won't do full damage but on your screen it looks like it was a direct hit. The problem is the positions of moving objects aren't the same on every machine at every moment and they can also be affected by lost packets or delayed packets as mentioned earlier. Where you see the enemy the server believes their position to be slightly different from that location. These same positional differences can also make combo explosions do partial damage or kill you even though you believed you weren't in their vicinity any longer.
One more thing that players have probably experienced that will hopefully drive home these differences in positions of things between the different machines involved in the multiplayer game: You swear you grabbed that enemy flag before they were able to capture yours or you were sure you made that flag return before the enemy picked it up. The server has the authority on who touched the flag first. The position of the enemy touching the flag before you just happened to be a little bit different on your screen than it was on the server at that moment and on the other guys screen it might not have even seemed close.
Hopefully, all of this is a good explanation of why a demo you take on your client machine isn't completely accurate on where players are when you see them in the demo and probably more importantly where you see them aiming isn't always accurate. The view rotation of the client you are spectating (where they are aiming) is client side you have to wait for that change to be transmitted to the server and then along to you before you will see the change in rotation on your end and this can also be affected by delays and lost packets. Even a demo on the server will see things different than what the client sees. The only way to see what the client sees, presuming no bugs and no tampering unlikely but probably possible, would be to see a demo that the client records himself.
With all of this in mind I hope I was able to convey the difficulty in detecting well hidden cheats. Later, I will apply some of this to the specific times mentioned in the video and point out what might mean something and what probably doesn't mean anything.
