<quotes clipped to focus on narrow point>
Specific to niche and protection -- D&D as a game, focuses on specific things at the expense of others. Being able to do certain things is exciting or not based on whether the rest of the game facilitates them.
In a super-low-magic homebrew, I had a character who knew how to
knap flint arrowheads, make beeswax waterproofing, weave his own cord and start fires with a bow-drill, and otherwise not just survive but stay a viable combat threat for years in the wild without resupply. In D&D, where even the more travel/survival-focused earlier editions usually amounted to checks to make sure you found the dungeon and travelled there in the least amount of time (minimizing purchased rations used and wandering monster checks made before getting there), those abilities just aren't going to see play (even if, as others have alluded to, there isn't a spell which can do it better).
So, to make a short answer long, yes -- a ranger's niche is weak in this game. Whether you want to protect it is another matter. And, as TiQuinn points out, if you do want a ranger-that-behaves-rangery to be a viable character type, there needs to be events in the game with differential outcomes based on whether there is a ranger there to tackle them.
If you want to do that, crafting a new ranger class to do so is a reasonable first step. However, where you will really need to start is both the skill section and the overall task resolution mechanisms in general.
Specific to the "army scout" idea -- this is a simplification, but IRL army scouts do a specific task well. They make sure they know where an opposing army is, as well as what other hazards and difficulties there are for their own wilderness-hardened (but less mobile) army behind them to move through the wilderness. In practical terms, they make sure that the eventual engagement occurs at the time and place most advantageous to their side, and with the highest information deficit in their favor (and their own side spent a minimal amount of resources getting into that position).
In D&D, for a proposed non-magical ranger, perhaps that translates to this: If the default is the DM saying that a wilderness encounter is happening and drawing out a battle map, party with a ranger gets to select from a set of pre-made maps. Depending on level and roll, they may also get placement of their choosing, deciding the direction their enemy is coming from, and eventually massive stealth bonuses for each character until they reveal themself (letting their opponents rush past and potentially get flanked or surrounded). It's still not going to be as effective as real life, in that being in optimal formation isn't quite as make-or-break in D&D as it is in real life*, high mobility (including flying) and poor zone-of-control means flanking already happens a lot, and the characters that most benefit from remaining hidden until the right time already can do so.
*and some things will never translate perfectly in a game where one successful hit doesn't drop or at least render ineffective most opponents.
Even then, this is only going to work when choosing the situation of the engagement is an option. In the common D&D scenario of invading a dungeon full on emplaced opponents, a ranger's abilities are going to be reduced to hopefully not being surprised by their opponents and not giving away their presence at inopportune times*. And that's reasonable, IRL -- much like cavalry, archery, etc., scouts are a unit type which can only be used most effectively in given scenarios. It's just really frustrating in a game where you generally play one character most of time if their applicability is too scenario specific. This is the same reason why 'still mostly effective when equipment-limited' types like monks and swashbucklers have generally been upgraded over the editions**.
*
although much like similar cinematic scenarios, that will depend on how well their comrades can successfully perform the 'don't touch anything' task.
**as that quality is recognized as not going to show up enough to make it more than a ribbon quality otherwise
So instead rangers generally get abilities which vaguely approximate (or at least allude to) doing what IRL army scout types would do (and then we argue over how appropriate an approximation they are). Many of the spells are kinda-sorta like choosing the optimal terrain situation, just without the pre-planning. If they weren't specifically coded as spells, most of them individually would be fine with me -- instead of casting
entangle, you maneuvered the enemy into an entrapping terrain feature no one had noticed until you invoked the ability. However, I would prefer something like a broad expansion of battlemaster-like maneuvers, or even better a broad expansion or using skills, items and terrain to your advantage (and then make rangers the best at doing so). But that will always run into the issue of making a class highly scenario--dependent.
Overall -- people have been doing good job of recommending other attempts at rangers which do not use spells. Most are, however, either a collection of abilities which 1) are not unlike the spells (just not magic) and 2) only vaguely approximate what IRL rangers do (and thus the eternal debate over applicability); or (like Level-Up) come from a game/version of the game which overall makes non-magical task resolution more interesting. If one of them speaks to you, go for it. I'm mostly just laying out the challenges of the task, not offering a specific solution (or poo-pooing the attempt, I want to make that clear -- I think it's a noble endeavor I love seeing).
