Have you checked that it didn't end up on the floor above or below? Gave ample time for the server to spawn the item? Sometimes takes a minute or two after you place it, most likely for security reasons against duping.

It's a clipping issue in first person, if you crouch and strafe you see gear clipping in front of your sight, making it worse, this clipping can also in some cases stop bullets.
It gets worse if you increase your FOV and when it's all black you are probably seeing the inside of the model.

Should add this seems to affect most things but are more noticeable with certain attachments and certain gear, hunters backpack for ex. seems more problematic than the normal hiking ones.

Non lethal weapons would be great, beanbags for shotgun, smoke/gas granades etc

It's true that RV also uses approximations but it's not as prominent as other fast game FPS.
If you for example look at Counter Strike Global Offensive you got a similar setup up only the client is allowed to approximate more. The server Tick/FPS on their normal servers are 64 and there max kb/s is 8K I think.

As the client isn't FPS limted by the server you can still have your 300FPS on the client only you often see bloodsplatter but don't get a hit. If a server runs with 128 Tick/FPS and I think it's 20K bandwidth for each client then this is much less an issue.

The difference with RV is that it wont let the client approximate as much, sure you still need some and I think Rocket has said he wants the servers to at least have a server FPS of 15.
You do need approximation for network games to limit the bandwidth usage etc, but if you compare RV to the source engine, a 64 tick CS:GO server doesn't limit your FPS but if you join a slow server with a bad connection on Arma your FPS takes a dip.

It's not a bug, it's the way the game works, just like most other games only it's more noticeable in Arma because of the massive maps.

The server needs to tell the client what it sees and what happens, the client then based on this renders the image. Games sometimes render based on approximations, Arma doesn't to this as much because it's a simulator at heart.
Most of the time you wont notice if a game uses guesswork to render, but in simulators its a big no no, the reason is because when you see a hit in Arma when you shoot you see it because the server sees it, not because your client guessed it.
This is the difference against Battlefield, CoD, CS that work on FPS/flow > accuracy, this is the part of their netcode everyone is screaming about, you shoot someone, you see blood, a hit, a clear shot but it doesn't hit, you miss because your client guessed and was wrong or you or someone guess and get it right but at the wrong time and someone shoots you even though your client says you gotten behind the wall etc.

The Real Virtuality engine doesn't work this way, it works on the opposite principle that accuracy > FPS/flow, this makes the client FPS bound somewhat more by the server FPS and the network communication speed. If your view includes more data needing to be verified by the server and taken in to account the limitations of the server and the network traffic will lower your FPS.

And even if you got a 1Gbit/sec fiber connection and playing on a super server with a super computer you are still limited by the netcode limitations which has to take in to account all the other players that might not have the same as you.
The idea is that every client should get a fairly identical amount of data and so on to make it an even playing field, a simulator has to show the same image to you and to your opponent or its not doing its job.

Blood tests aren't that rare, but they are small and can be hard to see, I've found a fair amount of them.

When you drink from a bottle you drink 2dl, also when you drink from the pond you presumably use your hand which gives you even less water each time.
A little water at a time has two perks against a system where you drink once and get full from thisty.
1: It's more realistic both in how people actually drink and how much time it takes.
2: It allows / requires you to drink regularly and take small fast swigs during breaks or downtime keeping your thirst at bay instead of you always having to wait to get thirsty to drink not to use water unnecessarily.

Would be better if quality could diminish and risk of breaking, for ex.

Pristine: 75% success, 25% success + being worn
Worn: 55% success, 45% success + being Damaged
Damaged: 35% success, 50% success + being Badly damaged, 15% success + being destroyed
Badly damaged: 15% success, 55% success + being worn, 15% success + being destroyed, 15% fail + being runied
Ruined: 5% success 65% success + being destroyed, 30% fail + being destroyed