As has been mentioned a few times upthread, the issue around the odds of failure depends on the particular game's baseline assumptions about what calls for a skill check and how the rules define expertise. I know that GURPS hasn't gotten a lot of love in this thread, but it's the game I'm most experienced with, so I'll use it as an example.
In GURPS, your base skill level "measures your odds of success at an 'average' task under adventuring conditions – in other words, in a stressful situation where the consequences of failure are significant" (GURPS Basic Set, p. 171). Under normal, non-adventuring circumstances, you roll with a +4 or higher bonus. (They include an airline pilot as an example on the same page.) As a result, even at relatively low point values, characters usually succeed at routine job tasks.
A recent character of mine, for example, is a security officer on a space navy vessel. She was built on 125 points and her base skill levels range from 10 to 14. Thus she's quite competent at her job. For routine tasks, she would roll with a bonus against a minimum of 14 (91% chance of success) for her weakest trained skills. For her core competencies, she has above a 98% chance of success. At that level, the GM—following the guidelines in the rules—assumes success unless there are high dramatic stakes for a critical success or failure. When we were in the setup phase for the adventure, she used a number of routine skills. The GM narrated the results, except for my edge skills. Even with those, she succeeded on nearly every check.
Later, however, we were in a battle with space pirates and there was a lot of time pressure (i.e., "adventuring conditions"). Under those circumstances, her skills were more modest, and had higher chances of failure. Still, though, she seemed quite dependable, especially at her best skills.
During a battle situation, she tried to use her blaster to hit a small control panel by a door from some distance. Her base skill is 14... generally considered to be at "expert" level. But, the smoke caused a -2 vision penalty. The size of the target and range led to another -3, meaning that her effective skill was a 9 (~38% chance of success). Now, she could have stopped to aim, which would have increased her chances significantly, but we were racing to get to another area, so she took the unaimed shot and missed. The chase was on. This was dramatically satisfying and fit with our sense of the fictional situation. She felt like a mid-range action hero.
If we wanted to make things even more cinematic, we would build the characters on higher point totals, leading to higher skill levels. Again, routine use would be a non-issue, but the characters would have better odds in more extreme circumstances: thicker smoke, longer range, smaller target, from a moving vehicle, etc.
While the D&D rules don't preclude any of this (I've played extensively in every edition since BECMI), in my experience it's easier to fine-tune character competency in a skill-based system.