Because even if they managed to include a Warlord that gave you the four things you mentioned in your list... someone else would have another list with even more Warlord stuff that they felt had to be in the class for it to be a Warlord. There are people out there who probably would only want a Warlord that exactly matched the 4E version (down to maneuver names and such) for the 5E Warlord to be valid.
So because there might possibly be a tiny handful of unhappy people, it's totally cool that nearly everyone who loved Warlords walked away unhappy with 5e? Now that's a hot take...
I just am unwilling to wipe the entirety of 4E-isms in 5E away just because they don't match.
Well, if you actually took my argument seriously, you'd know that I'm saying they don't match.
E.g., cantrips aren't at-wills with a twist or fresh paint. They reinforce caster dominance. They are actively fighting one of the most important purposes at-wills served, and there is no system which compensates for this. Or, for another example, hit dice aren't healing surges, and in fact serve literally the opposite purpose (extra healing on top of essentially-unlimited magical healing). You cannot even rework hit dice to make them work like surges (and the optional rule claiming to do so is, again, laughably missing the point by making them unbounded because there's no other way to get around the limitations of 5e hit dice healing.)
--------
So, building off the above and attempting to answer the initial question: Hit Dice
could have been an actually cross-edition modularity option. If they'd been handled differently, it
could have been easy to toggle between 3e-/5e-style "they're a bonus on top of unlimited magic healing," something relatively 1e-style where healing is slow and difficult (e.g. maybe you only recover
one hit die per long rest, so they're precious), or something 4e-style where they're beefier but act as a cap on healing. Likewise, it wouldn't have been THAT hard to, for example, make a list of Combat Stunts that use up one attack roll to do something cool
and do damage (e.g., stuff like Shove, but actually, y'know,
interesting and making-fights-end-sooner rather than delaying-when-the-fight-ends), with toggles so that the 2e approach requires you to spend Weapon Proficiencies (and giving tables of such proficiencies for various classes), a 3e approach that locks them behind feats, and a 4e approach that lets non-cantrip classes (e.g. Paladin, Fighter, Barbarian, Rogue, etc.) pick some up, perhaps on a schedule similar to cantrips for full casters, e.g. Paladins get fewer because they're part-spellcaster, while Fighters get the most. Etc. Skills could have toggles to 2e-style NWPs, 3e-style point-based skills, or 4e-style broad skills and skill challenges. Etc.
THAT is what modularity would look like. Multi-state toggles carefully constructed to assist the
actual spirit of the rules of various prior editions, rather than superficial mimicry.
And I would absolutely include "zero level" rules as a similar component, a toggle for groups that like really slow progression and really legit actually
zero-to-hero play. Likewise, learning from 13th Age, rules for "incremental advancements" so groups can smooth out some of the chunky aspect of levelling up
and feel like they're making progress while still keeping the pace relatively slow.