Heroes 5.5 WarGame Edition - Tactical Edition.

The variability from competing perks is much stronger than the variability of "anything goes." Because it forces you to think more carefully about how you develop your character.
You choose one, and you lose the other.
And with you, it turns out: you choose one, but you'll think about the second one later.
This is a very simple choice. The game is not just about different combinations.
In this case, the combinations will only win in single-player, but the mod is for multiplayer ;)

Okay, now I understand your idea. For me, this kind of choice usually starts more often from level 2 than from level 1. For example, I took Evasion – formally, I didn't lose anything. But as soon as I took Resilience or Readiness, I lost the ability to upgrade Fire Resistance or Stone Skin. This allows me to take Evasion when Evasion is really needed, and not just to develop some rigid branch or give up on others.
In my opinion, the factor of variability is stronger than your argument. That is, 10 options are better than 2-4 linear ones, the choice of which is rigidly defined already at the first pick.

In fact, choosing a perk at level 1 already has a significant impact. The fact that you take Evasion suggests that the next picks may offer Resilience or Readiness, which reduces the chances of taking other perks at levels 2-3 if they were planned. Under these conditions, it is better not to take Evasion without a real need (and prefer basic perks with access to 1 perk at level 2 or even dead-end perks without access). A consequence of your approach is also that there are no basic perks with zero access (dead-end) - it is impossible to take a "safe" perk at level 1 that does not spoil the probabilities of upgrading other perks at levels 2-3.

The binary scheme is very inconvenient in design, as it forces you to introduce too many restrictions that did not exist in the ternary one. I admit that your specific scheme turned out to be a very successful implementation under such limited conditions.

Added 13 minutes ago

There is no alternative, upgrade a tent or a ballista. Rune Machines will enhance both.

Why shouldn't the player have the right to combine?
He can upgrade medicine + ballistics in different ways, artillery + ballistics, artillery + medicine. This choice (combinations) is much wider than just choosing 1 out of 3.
That is, with a choice of 2 out of 3, formally there are only 3 directions in both cases. But with a binary scheme, there is no internal variability. In my scheme, I can upgrade medicine + ballistics in 3 different ways: with plague, with crafting, and with rapid fire. And by the way, there is also a 4th direction, which is to upgrade all 3 (medicine, ballistics, and artillery). And that will also work well.

Let's consider only engineering as an example.
I will list my options (all of them are effective):


1) Ballistics + Rapid Fire, Military Medicine (strong ballista + healing)
2) Ballistics + Rapid Fire, Artillery (strong ballista + artillery barrage, if there is no tent or damage is more important than healing)
3) Military Medicine + Military Crafting, Ballistics (strong healing + ballista)
4) Military Medicine + Military Crafting, Artillery (strong healing + artillery barrage, if there is no ballista or weak stats)
5) Military Medicine + Plague Tent, Ballistics (healing/barrage + ballista, flexibility: the ability to switch between the strongest barrage without healing and barrage with healing)
6) Military Medicine + Plague Tent, Artillery (healing/barrage + artillery barrage, the ability to provide both healing and barrage in the absence of a ballista or weak stats)
7) Military Medicine + Military Crafting + Plague Tent (strongest healing or good barrage)
8) Artillery + Sulfur Cloud + Ballistics (strong artillery barrage + ballista, if damage is needed, but the stats for the ballista are not top-notch)
9) Artillery + Sulfur Cloud + Military Medicine (strong artillery barrage + healing, if there is no ballista or the stats do not allow you to play)
10) Artillery + Sulfur Cloud + Fire Arrows (strong artillery barrage + strong unguided ballista, less dependent on stats)
11) Ballistics + Artillery + Military Medicine (versatility, if there is enough healing, and the ballista is not good enough to develop to rapid fire)

Options according to your scheme:


1) Rune Machines + Rapid Fire + Ballista
2) Rune Machines + Rapid Fire + Tent
3) Rune Machines + Tent + Plague Tent
4) Rune Machines + Tent + Machine Devotion
5) Rune Machines + Rapid Fire + Machine Devotion
6) Machine Devotion + Catapult + Rune Machines
7) Machine Devotion + Catapult + Sulfur Cloud

How many of them will work effectively?