No it's realistic. People can be stabilized from death in under 6 seconds from any kind of wounds.
But is it realistic in a pseudo-medieval context? Would a medieval-era chirurgeon have been able to stabilize a dying person in under 6 seconds? Regardless of era, is it realistic for the person doing the stabilizing to be able to do it without any medical equipment and regardless of
why the person is dying?
Like, how do you save someone who's been poisoned without some sort of antitoxin? How do you save someone who's suffered third-degree burns with nothing but your bare hands? How do you save someone who's been melted by acid without any form of medical equipment? How do you save someone who's been frozen without a thermal blanket at the very least?
My issue with the rules isn't just the time span but the fact that, in 5e at least, you can stop someone from dying, regardless of what put them in that state, without needing to actually do anything within the fiction of the game. A flat DC 10 Medicine check is all that's needed. It's a 100% gamist mechanic with no simulationist backing whatsoever.
The more I think about this issue, the more I'm inclined to house rule non-magical healing. Currently in 5e, you can either make a DC 10 Medicine check to stabilize a dying creature or do so automatically by spending one use of a healer's kit (or by casting
spare the dying). I might combine the two - so you need to expend one use of a healer's kit
and make a DC 10 Medicine check. I wonder if it should also be like Concentration and be DC 10 or half the damage taken, whichever is higher. So someone who's taken massive damage would be harder to stabilize. But then the harder you make non-magical stabilization, the more players will just gravitate towards magical healing instead. Something to discuss with my players, I think.