Yeah, my projects all the time have little instances that crop up where I'm like "Meh, that's kinda lame. Oh well, no one will try doing THAT"
Usually a week or so later I get off my butt and get rid of the problem (sometimes) because well, I'm not too bright with game design, so if -I- notice a flaw, probably everyone will notice it too.
Case in point:
In my space-ship survival game, a glitch in the bullet programming makes one particular enemy fire a bullet directly behind it. The player could conceivably guide that bullet all the way to the end of the stage, hit the "End Level" marker in line with the bullet, and have the bullet kill the player while the "CONGRATURATIONS" screen is playing, fucking up the game (as the player would respawn with the Victory Screen covering up everything, making the level virtually unwinnable until a Game Over).
I saw that, and I'm like "Nah, no one is ever going to do that"
But a little later, I realized that someone is very well going to do that, so fixed up the coding to destroy all bullets when the "End Level" is touched.
Once again common sense saves the day!