Do you have some examples of "fluff-thin grid-filler classes"?
I was gonna mention Seeker and Rune Priest, but then I remember those were doubling down on combination already covered... The Battlemind was pushing it too. By PHB3 they really were reaching for some more content... But the races were great!
Warden was a Primal Defender and it was awesome! Lots of cool self-transformation, lots of plant flavor, and they made the ground around them difficult terrain! They could use either CON or WIS as AC stat so you could have a really solid guy, or a wise warrior in tune with nature.
That said, I think you could do the same concept in 5e with a Barbarian subclass that gains special transformations when they Rage, since the Warden's forms were dailies. It's a bit Beast Barbarian but it was more Shamanic with forms such as 'Form of the Winter Herald', 'Form of the Thunderbird', 'Form of the Laughing Killer' and so forth and not just "you get claws". I worked on that concept for a while so maybe I should revisit it? I'd probably do it as basically replacement for the basic Rage features, otherwise it would be too strong... So you can pick your standard Rage or your special rages depending on the situation.
In comparison your skill expert fighter will seem like a dum-dum.
Your fighter is already a dum-dum.
oes each class need exactly one primary and one secondary, or are some class ideas naturally flexible enough that they could choose from among several things as to what they were best and second best at, or maybe have a primary and the two in third place with no second place? Do folks talking about power sources fall in to similar traps where they need to pigeonhole things and make it hard to get a Witch or Bard that crosses the now rigidly imposed spell list lines? If a Gish is the mythically exact 50/50 between classes, is there an exact 50/50 between roles, or is primary/secondary role an immutable classification system?
If you try to make a class that can be everything you end up with the 5e Monk. Limiting the roles of a class in the
first release makes it easier for you to know what you can break when you expand the class later. You basically place barriers just to see where people will run into them so you know where to make holes in your barriers. The other barriers are still there, but since people can just go through the holes they don' t notice.
The Warlord concept is perfectly fulfilled by the 5e Battlemaster with appropriate maneuvers. He can do everything people want a Warlord to be able to do. That's conceptually. (Grant attacks, provide temp hp, reposition allies, grant an ally advantage, even push enemies away).
The mechanical implementation is where this really falls short for most people. They want to be able to grant more attacks, grant more temp hp, reposition more allies, push more enemies away.
Can such mechanics be made into a class. Almost certainly. Should they be made into a class is the real question though. Can a class focused on those things even more than a battlemaster meet balance requirements, can a class based on them add to the conceptual space of 5e in any way, will adding this class in take away too much from other already existing classes, etc?
IMO. Part of the beauty of the subclass design space is you don't have to worry about any of that.
I guess we don't need a Wizard if we have the Eldritch Knight?
"College Professor" would hire a fencing coach before adventuring.
They wouldn't be a fighter though. A fighter is an elite warrior.
A lot of people seem to forget the 'elite warrior' part of the Fighter. For a lot of people, the Fighter taking their attack action is the 'baseline' competency in combat. Isn't it weird? To me, at least, the baseline should be the Valor Bard swinging a sword or the Rogue without Sneak Attack.
The problem is that anything left to mechanically distinguish a Warlord for 5e would not be something the 4e Warlord had.
I'm totally fine with a 5e Warlord having little to do, mechanically, with the 4e Warlord, as long they can support their allies and do it every turn if they so wish.
Speaking of beastmaster, I love the concept but the issue is that it is kinda hard to bake 'has a sidekick' into the class' rules. Now some subclasses have done it better but they still are a bit weird, but others tending to be some sort of magical summons makes their odd behaviour a bit more tolerable. But like can't my druid just awaken a bear and befriend them and have them adventure with them? Can't my paladin have a squire or can't any class just hire a sidekick or buy a trained animal? And obviously they can and those sidekicks tend to be better than the beastmaster's companion, so why shouldn't the ranger just play some other subclass and get a sidekick by other means?
However, I've also seen (with other groups) players react poorly to a player wanting to bring in an XP drain. I've seen players reluctant to accept companions because they would be an XP drain.
Baking the companion into the class circumvents these issues, since it will scale and won't take XP from the party. Granted, it can still have issues since a poorly designed companion feature may not scale sufficiently well, but that's more a function of bad implemention. Just because something can/has been done poorly doesn't mean it can't be done well.
On the subject of Beastmaster, I've been saying for a while that there should just be a 'Beast' class and the player who wants to play a Beastmaster just brings the Beast PC along and controls both. Then the DM just adjusts their encounter in the same way they would if another player had joined the group.
Then, you give the Ranger and/or Fighter special features that make teamup maneuvers possible and make the Beast PC also give bonuses in similar circumstances so they have synergy but not all of the main PC's subclass features need to be dedicated to balancing the sidekick.
