I think this is one of the bigger complaints I have with some GMs. If preparing for the known unknowns (possibilities of bandit ambuses) just means we're going to get ambushed by bears instead (an unknown unknown), what's the point? Sure we can't expect to have perfect knowledge of what's before us, but if we have reasonable expectations that the danger we need to concern ourselves with is bandit attacks, doesn't that seem to imply that bear attacks are not a reasonable expection?
In a table I had to walk away from the DM operated like this. We could prepare for rain...and we'd get a tornado. We could prepare for sun, and we'd get a blizzard. The consequences of us attempting to prepare for reasonable possible outcomes was more extreme impossible outcomes, things we could never have seen coming.
Replying to both you and Von's last post...
i do not think what was attributed with the bugbear bandit case was what was being described by myself at all. nor is the "rain" vs tornado.
My case invoked more of random monster encounters checks for moving slowly, taking more time, literally spending more time out in the area. I have not seen linking how many random monmster checks to how long you are out where they can happen as a problem in general. i also linked moving slowing and cautiously when pursuing a foe as leading to delaying the point at which you catch the group you are chasing... again something not "out of the blue."
I would also rule that moving to travel at the top of a ridge or whatever for better viewpoints as having its own drawbacks as well, like possibly being easier to spot (if you have moved out from the natural cover of "light foliage" or if you are now putting yourself at a distance you would need to cover to get back to the ground if something happened (assuming the whole party, horses and all, did not go to the ridge with you) or at the very least from the ridge your ability to spot tracks and clues from the groups trail would be less. That spearated some on ridge and some not might well be an issue if one group gets a wandering encounter.
All of these are allowing the player's choices to matter as far as how their characters PRIORITIZE the risks - known and unknoiwn.
Consider, what if the bad guys had dropped caltrops along their path? Then the ridge move would bypass those for all those choosing the ridge even though it was an unknown threat. Similarly, there might be varmints on the ridge
But the key is this - in these cases the players are choosing to prioritize some risks over others, priritize some objectives over others.
"We can handle any small ambush but if the main group gets too close to their base we will be screwed completely, so lets risk the ambush and hurry."
That is a net trade-off where the objectives are influencing the determination of which risk to take.
"We need to not give ourselves away so running into an ambush is out of the question. if they get to the base, we have enough stealth to sneak in anyway as long as we surprise them and an ambush blows that. . So lets take this slow and use surprise rather than speed."
Again, same situation but different priorities based on different plans for how to deal with the challenges.
In either case, the choices lead to advantages and disadvantages determined by their nature... not some newly created GM screw job where the PCs have to be shown wrong.
its simply using the already established principle that you can gain a bonus for a given thing if you give up something else that is related ans sensible - in a broad sense - as it appears all through the 5e rules (and many rules in many games.)
Obviously, sometimes circumstances themselves would apply advantage- like in this case of pursuing a significantly large force the tracking would be against a very easy DC due to the size of the enemy and their rate/nature of travel. (had they slowed down to cover tracks, reduce signs of passage - that could change that but at a price.)
but hey, thats me. Obviously other GMs can rule differently.