Jester David
Hero
Ultimately, though, whether you expend a resource for it or not, an action-grant that costs the same action is pretty neatly balanced. Obviously so, in terms of the action economy. But, even in terms of doubling-down on whatever the best action that round may be, it comes out more flexible, but less potent, than simply adding a second character able to take valuable actions, himself. For instance, D&D traditionally had a spell/round limit (3e broke with that tradition, obviously), so granting a caster an action that can be used for spellcasting (far more powerful than granting an attack or cantrip, obviously), seems as powerful and game-breaking as that forbidden 2/spells round. Yet, if instead of an action-granting character, you simply added a second caster, you'd have the same 2/spells round, /and/ twice as many slots.
So, no, action-granting, even action granting far in excess of what the Warlord traditionally was able to do, isn't inherently broken in 5e. 5e's balance is simply looser. It's hard to imagine anything a Warlord might do that could be any worse than what casters already can.
It's not particularly balanced.I don't see an issue granting spells.
If you had 2 wizards, you could cast 2 spells, with 2x the slots.
If you have a wizard and warlord, you could cast 2 spells, but only 1x the slots.
Or to twist it a bit, "you can take a spell slot from a willing creature to cast a spell".
It has it's uses, but it's not an ability your going to be spamming a lot. Maybe if the cleric's bless drops you can renew it for them, or if the enemy is in perfect fireball position, but you'll quickly burn out of slots with 2 people casting.
It's the equivalent of giving one character twice as many actions. Which seems fine when you consider they're replacing another character in combat. However, the replaced character would still have all their potential non-combat options. So it's everything the one character can do and everything else.
Think about it this way, you could do that right now. Build a character where every feat and class option was based around non-combat options or bonus actions, and then just let another character act during combat.
Would you allow a player to run two characters at once: one combat and one RP/exploration?
Additionally, it's replacing an ineffective turn with an effective turn. Not all characters are effective each round. The melee characters outside of melee range, the ranged characters in melee, the pyromancer facing a fire elemental, the rogue without a flank buddy, etc. In many situations in a dynamic game, a character's turn is not 100% effective. However, a warlord with at-will action granting is always effective, since they can choose who takes the additional action. Each turn is extra optimized, and does far more than the character could do themselves with their own action. They can effectively cast spells as as a sorcerer, attack as often as a fighter, strike as hard as a rogue, heal as good as a cleric, and more. Whatever is needed any given round. It's incredibly flexible and thus incredibly powerful.
Even in 4e this was an issue. The lazylord was not a standard build, and was a fan build and one not really viewed seriously at first (as seen by its original name as "the princess warlord"). The vast majority of warlord powers involved attacking. The lazylord was problematic as it could dump stat the warlord's attack stats and focus on the presumed kicker stats for superiour bonuses. It was a great munchkin build, being very effective.
It's also very much not what I'd base a 5e warlord on, since it's a message board build. It's very, very unlikely to be how the majority of warlord players designed their characters.