Marandahir
Crown-Forester (he/him)
I will say something a little controversial.
There will always be a weakest class because some classes are geared to aspects of the game that is not emphasized or apparent at all tables.
This is a big part of it.
If we all played with all the same emphasis on all the same rules - i.e., everyone cared significantly about rations and carry weight and daily march and sleep hours and exploration tracking - then the Ranger as seen in the PHB may actually function significantly better. Same thing with the Berserker Barbarian - if exhaustion is a risk for EVERYONE, then the Barbarian getting a level of exhaustion isn't as much a detriment because everyone will want to take that time to retreat from the dungeon and rest until healed.
The thing is, this is a modular game, with a lot of different focuses at different tables, and the classes provided need to be a constant, so that they can keep up regardless of the DM or the campaign style. That doesn't mean the base rules should expect a skipping of Session Zero (If it's an Ice Age-style game and I come prepared with my Hot-desert focused Ranger, we've got a problem of expectations between DM and the player). But it DOES mean that if the campaign takes a sharp left from 5 levels of fighting Ogres to 15 levels of fighting undead, that Favored Enemy: Giants feature is now useless (as are the necromancy-focused Shadow Sorcerer and the Evil Cleric who learned Inflict Wounds rather than Cure Wounds).
Retooling can and should be a base option, and that UA article gives each player class that didn't already have some measure of retooling a way to do it (and some to do it in capacities that they didn't have previously).
Ultimately, though, a more general-use feature is probably the simpler solution. Playing as Link being able to climb cliffs and explore and run with special stamina features is a a very Rangery replacement for favored terrains. We're looking more at the generalist survivalist rather than the specialised one. It doesn't mean it's right for ALL Ranger characters, but having opt-in options for each class is not a bad thing.
Now, we could talk about trying to make fixes and bonus features and how that causes power creep, but if the creep is across the board, then we just need to rebalance the adventures. Throw in an extra goblin or two to the encounter once you realise that the players are lasting longer than they used to.
In this thread, I also think there's a fundamental disconnect between people who want the player to play the Beastmaster Ranger the specific way that the mechanics suggest, versus people who want the player to have the option to play the Beastmaster Ranger the specific way that the the player interprets the flavor to suggest. These are not necessarily the same.
In past editions, some classes had very specific flavor, and others were broad-tent flavour. In my understanding and experience with 5e these last 6 years, the current edition tries to give as big a flavour tent as possible to each class without stepping on the other classes' toes (mixing several metaphors, I know). That doesn't mean that subclasses from different classes might feel very adjacently related - there's story overlap between the Oath of the Watchers Paladin and the Horizon Walker Ranger by intent (see D&D Beyond interview on the Paladin's UA article). I'd surmise this is the case for many subclasses - Oath of the Ancients Paladin is one way to approach the old 4e Warden concept, as well as approaching some types of past edition Ranger concepts. But you can approach those with Barbarian and Ranger as well. There isn't a one-to-one conversion from past editions' classes and archetypes, because they've been reimagined in a 5e concept. The idea is to ideally be able to tell the same stories, and to tell new stories, and to tell stories that appear in other popular (esp. fantasy) fiction stories.
As an example, with the 5e PHB, I couldn't adequately tell the story of a Druid and her animal companion, as seen as a default option in 3.5e's PHB. Variant class features UA provides me some semblance of that option without locking me into a specific subclass (Wild Companion feature, enhancing Wild Shape). It's still a familiar, not an animal companion like the Beastmaster Ranger, but it speaks to the concept in a way I couldn't tell previously. Mielikki forbid having to own a figurine of wondrous power to represent past-edition class features come the edition shift.
With the Beastmaster Ranger, then, the issue comes that there are many players who want to play a Boy-and-His-Dog story for their character (or maybe a Girl-and-Her-Giant-Frog, ymmv). This is a very different approach from players who want to play an animal-tamer who uses the wolf until it dies and then starts training a lion or a bear until it dies, etc. And it's different from the grizzled woodwose who is a friend to all nature but not so much to other humanoids, and befriends animals as they come and go, swapping them out as they travel.
The question is, SHOULD the Beastmaster Ranger encompass all of these character archetypes and more, or should it represent only one or some of them? Are subclasses narrow narrative slivers, a representative of a specific character archetype in fiction, or are they a broader piece of the puzzle here? I'd argue that some are more specific and some are less so, often relating to the specificity of the parent class (Fighter is a heck-more broad than Warlock, for example). So is the Beastmaster Ranger supposed to be ONE way of telling the animal companion Ranger character of past editions, or is it supposed to represent many ways? At what line do we draw the difference between "these mechanics don't exactly replicate my class feature, primal evocation, and martial exploit choices from 4e, so it can't translate my Ranger" and "these mechanics don't exactly represent what I could do last edition, but they respect the character I was playing?"
The same question can be said to new characters, and player expectation. Where do we draw the line between creating features to mold around player-desired abilities vs creating features that mold the way players play their characters? This isn't make-believe; we're not making it all up on the spot, and there are acceptable limits to player expectation going in - the constraints help develop the character in ways the player might not have anticipated. But they also can be so constrained that the character they might have played no longer can exist, because the mechanics keep pushing them to act contrary to their character's intentions and motives.
If you couldn't tell, I believe that the UA Ranger in Variant Class Features MOSTLY fixes the problems of the narrative, especially if you treat it as opt-in features, rather than full-scale replacement by the DM in the campaign's dossier of house-rules. But I'm not opposed to a variant subclass that trades off versatility of animal companion (and perhaps power) to have a pet that grows and evolves with the player, a la a Starter Pokémon. I think there's room for this sort of feature in the game with weapons of legacy too, though we haven't seen it fully implemented in this edition. I'm not sure what these implementations would look like. I also believe that the easiest solution is using something already made or partially made (as is the case for UA) and tweaking as we go if the power disparity is too high or too low for the campaign.
As many at WotC have said, it's almost always better to playtest new concepts without worrying about if they're broken on the high end, so as to get the "does this feel fun" and "does this respect what I want to do with the character" pillars of design right. After that, the restrictions can be figured out as they become apparent.
I think we're seeing in this forum a very small minority of players who feel that the Ranger is just fine as it is in the PHB. That isn't to say there isn't a majority of players who feel this way, but rather, there's significantly enough respondents to WotC's big data research on the game that felt the class is underpowered, that it's worth making opt-in revisions to the class. I would suspect the vast majority of respondents are NOT white-board forum theorycrafters, but rather players with experience with the class. Most of them still said the class was fine. But a lot of them said it wasn't, and the sort of data they draw in on those surveys, it's far more than can be attributed to forum posters running spreadsheets on who's more powerful here or there.
Most of us are just frustrated that our character isn't working right for the campaign, despite our best efforts and intentions. The game can't fix the campaign, but it can tinker with the core rules to more adequately reflect the base assumptions of broad-tent game play.