I try to avoid miniatures whenever possible, for two reasons. First, because it slows down play. And second, because it tends to move the imagination of the player from the first person perspective to the third person. The player ceases to imagine the situation that they are in, and instead just thinks in terms of the too simple tokens and maps that reference that situation.
I agree completely. But, I can't ignore the obvious enjoyment it brings out in others...so it is a balancing act that I manage group to group.
I agree that it is important to tell the players what they see, hear, and smell, but I draw the line at telling them what they feel. It's not my job to play the PC. If the PC believes his character would handle the hardship stoically, or the PC believes his character finds the time in the jungle refreshing and beautiful, it's not my job to tell the PC otherwise.
So I would rewrite your lede to remove all direct emotional content, and say the same thing something like this:
How the players are supposed to emotionally respond to this or characterize their response to this is up to the player and their imagination. Ideally, I'd have players that can respond to my lede, but I don't at the moment have a particularly thespian group, so that part of play isn't particular in the foreground at least with this group.
As a player, I find it jarring when the DM takes a directorial stance and tells me my motivations and feelings in the scene.
Yes, I see what you mean...and even I have been jarred in a similar fashion. Though, as a GM, I welcome a player asserting, or reasserting, their players reaction because it speaks to a higher degree of investment in the game (and I'm able to learn how to communicate better with the player).
I look at emotion as a subtle tool of manipulation rather than an authoritative stance. People are very used to hearing judgment in simple expository (unfortunately so) and tend to let it wash by without comment. It goes against every fiber in my body to use loaded emotion and cliche in my writing but I can't help notice the success it brings during role-playing sessions. It becomes a tool for motivation. By hinting at emotion, or possible emotion (players are free to correct you), the GM can lure, hide, agitate, pacify or even confound a player's attitude toward a scene in a way that a descriptive narrative tends to fall short. And that is really what I think I'm edging toward: the ability to keep things moving, getting the players invested and eliciting action more than waiting for it. I guess emotion has become my clumsy shorthand...but, like I said, perhaps I'm just too lazy.
As for vagueness, the easiest thing in the world is to be vague. While it's true that vagueness allows for flexibility, the danger in not being concrete is that IME the more likely outcome is that anything left unsaid doesn't exist. If the floor doesn't start out rocky and uneven, with a marked slope to the east, then the floor will be featureless. Even worse, there can be a dangerous trap that DMs can inadvertently introduce whereby they discourage players from asking questions or investigating the environment. The most succinct example is the trap on the door that didn't exist until the player choose to search for it. There is a tendency among DMs that are vague to bias their in game detailing toward creating conflict and complications, with the result that the wise metagamer avoids providing the DM fodder by exploring or interacting with the environment. The last thing I want to do is get into a situation where I'm punishing the players exploration of the shared imaginary space.
Yes, it is easy to be vague. Perhaps I was being a bit too glib. The lack of precise expository detail is needed for a couple reasons and should be paired with a couple of other tools. First, I think brevity is important to keep games flowing. There was a time when I typed out or even borrowed long narratives describing my world, my NPCs, my locations or even my politics. Even the greatest orators have to watch peoples eyes glaze over after the first few seconds. This can be dealt with through cadence, tone, posture or even volume; but, in a game, I have found interaction far exceeds these tools. If you have a large amount of narrative that 'needs' to be communicated (this need has completely disappeared for me) you can create breaks with questions, create dialogue or even use props. Second, as I stated before (and you are worried about) it allows a type of co-creation between the GM and the players (losing a bit of control is also something I have gravitated toward as I age) that, I feel, invests the players more...assuming they ask questions. Which is the crux of it. In order for this to work motivation has to be established.
So, I pair my lack of detail with emotional undertones and, whenever I can, a sense of urgency. This doesn't have to be straight conflict. These types of situations can be created by simply asking each player what their character is feeling or thinking as opposed to what their character is doing. A quick summation of what the players have stated (a surprisingly effective tool) another, perhaps more acute, description and then off toward character actions. Don't underestimate the power of a weak emotional descriptor when speaking to a group of people...it is surprisingly effective.
As for the meta gaming, yes, every GM will have to hone their chops. I distinctly remember being on guard whenever a room had more than a sentence of description or being wary of any NPC that had more than a few lines; but, these are easily addressed if one is willing to spend the time. Hell, I used to go out of my way to create lovable NPCs just so I could kill them for story motivation...pretty blunt (though it has been quite effective for a couple of authors now). I think this is in the area of a tool in the hands of a beginner as opposed to a master: we all have to spend the time if we want to get close to that master level.