The only problem I have with take 10 mechanics is that there's a big jump between "Auto success" and "Fail more often than succeed" that this generates. There's no "Slight risk of failure" that might make the players consider their options, or at least make trade-off decisions about risk versus speed/resources used interesting.
However, you can still make auto-success situations interesting, because there's still an action cost to performing a skill. When trapped in a room, and needing to escape, should the rogue engage in battle, or investigate the door? Should the rogue spend an extra action checking for traps? Even if both those were single action and auto-success, it could still be nail-biting for the player team.