There isn’t that much thinking going on, but for understandable reasons (at least for me). It all runs down to having to make their decisions in a split-second, and likely relying on reactions and reflexes more than conscious decision-making.
Those people don’t have all the info that you have. In particular, they don’t necessarily know how long the target has been sleeping. Thus, to them, they might still sleep for a couple of seconds more, or they might be waking up right now. Even if they did have that piece of info, they might estimate the timing wrong, or slightly panic and forget the timing.
They may still be reacting to the same cues that caused you to sleep the target - an enraged scientist jumping in, or a burly omnic rushing past. Or they might recognize a disabled enemy as a free target, and only realize afterwards which target it was (and that they were slept, not just stunned or knocked down) - too late to not shoot it, early enough to immediately regret it.
These may happen even if in your POV they’re coming leisurely from spawn. They may be rushing in, coming into a firefight, paying attention to many other things than that exact situation. Thus, their learned default reactions (dangerous target -> shoot it) take over, even if they’re not the “correct” ones this time.
All this is annoying and hard to override, but can be learned out of. It’s a long way there, though, and it’s not something you can realistically help with - it’s their business to get better at their game.
It’s frustrating, I know - I’ve been on all sides (sleeper, sleepee, awakener) of that situation before. So is there anything you (as Ana) could do to help this? Well, yeah, though not everyone likes the solution and it doesn’t solve every situation. While you can’t control what people will do with a certain piece of info, you can help them get more relevant intel. So if you’re in voice, call out your key actions. Calls like “Sleeping BOB”, “Hog 'naded”, “Genji, nano in two, one, go” can go a surprisingly long way.
