2 reps in reserve is fine and far less painful, but you need to go to actual failure often enough to know where failure is on each set. I’m nerdy enough to suggest rolling a 20 sided die for each set, and on a 1 take it to failure it’s not that complicated and keeps your predictions honest.
As I understand it taking a set near failure works reasonably anywhere between 5 to 30 reps, but 30 well controlled reps with good form * 3+ sets for each muscle group gets really boring.
Boring is subjective though. For some like me the ideal weight gives endorphins where as too much feels like cortisol. Too light is sort of nothing. So I aim for that "yeah I pushed something" feeling. Which isn't failure.
Let’s be realistic, everyone goes through periods when they just don’t want to work out.
So optimal in terms of personal preference is defiantly worth considering alongside optimal in terms of results, but optimal in terms of returns on effort defiantly has a place at some point in our lives.
If you actually can’t use proper form that’s already failure of one of the muscles you’re trying to train. However many people resort to improper form well before that point.
Further the risk isn’t just injury, using excessive weights you can’t properly train with can mean failure to provide proper stimulus to a muscle you’re targeting.
As I understand it taking a set near failure works reasonably anywhere between 5 to 30 reps, but 30 well controlled reps with good form * 3+ sets for each muscle group gets really boring.