By that definition the real world is a railroad. Is that really what you mean to say?
Putting theological conundrums to one side, the real world is not a game, and the causal forces in the real world are
actual causal forces, not imagined proxies for someone's authorial decisions.
Being captured by pursuers because one reached a dead end is something that sometimes happens in the real world, due to the way the world is.
Being captured by pursuers because one reached a dead end might also be something that happens in a RPG. But that is not because of "objective" causal forces. It's because someone, via some process, established that there is no secret door in the blank stone wall. We're discussing different forms that process might take.
In the fiction, there are three possible independent outcomes on a secret door search:
1. A door is found.
2. A door is not found because while a door is present the search was for whatever reason unsuccessful.
3. A door is not found because there is no door there to find. (no matter how good your search is, i.e. no matter what the die says)
In the fiction, the PC should have no way of knowing whether a failure is due to 2 or 3 above; and thus neither should the player at the table.
Now in your system you'll probably say 3 can't happen
3 can happen.
Possible failure narrations could include, in no particular order, any of the following (depending on what the GM thinks up and how it seems to fit the unfolding situation):
(A) You start to search for a door - but then it finds you before you find it! A secret door opens and a squad of guards comes through it. It's the pursuers on one side and this squad on the other! - what do you do?
(B) You search for a door, but there seems to be nothing there. Or, at least, nothing you find before you hear the sounds of closing pursuers. What do you do?
(C) As you search desperately for a door, your pursuers catch up to you. The one in the lead mocks you: "If you'd done your homework, you'd know there are no secret ways in or out of this fortress!" What do you do?
(B) and (C) are both consistent with your (3) as well as your (2), and (C) makes (3) more likely than (2), assuming the lead pursuer is a reliable source of information about the fortress.
Too much likelihood of finding secret doors in every wall even where they don't make sense?
Why wouldn't it make sense that a bare stone wall in a D&D-type building has a secret door in it. They're pretty standard architectural features!
As for too many - I just don't think it's going to come up that often. Unless you're playing Gosford Park - the RPG, and then secret doors/passages/priest holes/maid creeps
should be pretty common, shouldn't they?
The players, sometimes via their dice and other times not, are forcing the DM's narration into saying what they want it to say. In other words, the players are railroading the DM until and unless a roll fails; at which point the DM can have some input.
This is the first time I've heard
player success in action resolution described as "railroading the GM". So when the players kill all the orcs you put into your dungeon, they're railroading you? Because you have to accept the outcome of the combat mechanics?