Grando
Auto-translated
Overall, you've written everything correctly. And by the way, this implies that the number of Elves can be chosen without using a mathematical evaluation method, but rather based on intuition. By asking (and carefully listening to!) the opinions of top players. Even at the level of feeling – this is currently too powerful in the game (it strongly affects the final result), and this is useless. These will be guidelines for which direction the numbers should change from what they are now.
Regarding Warlords – I see the problem there as the project's overall development having stalled. They would have continued experimenting, even with Brogh, but it became clear that the project's engine had shut down, and the community's morale was steadily declining.
However, it's a huge step forward compared to SoD, which is obvious to me. And the changes that have already been made, including minor balance adjustments and starting armies, are simply excellent.
When I compare Warlords and HotA, I mean some kind of overall cumulative desire to do everything in terms of balance in a smart way. In Warlords, they were ready to reconsider anything. Any immutable foundations for the sake of balance and overcoming excessive randomness in the game (so that skill, in its original meaning, would decide the outcome, and not the skill that Reddit has). For example, I recall that they were considering shuffling spells from one school of magic to another, and from one level to another. A bank, credits (insurance against the generator simply not placing certain units), monster experience (increasing the importance of micro, the ability to fight with minimal losses), and by the way, you can play with these features or without them, there are checkboxes. The project fizzled out, but the initial desire to balance the game and increase the weight of skill over randomness (which I call "mathematics") was higher.
I'm not criticizing now. The main idea of making HotA more popular, more mainstream, is clear. But the consequences of this are exactly what I'm writing about. When the question arises – to make the balance less precise but simpler, more understandable, prettier, without unnecessary checkboxes. Or to make it more complex, but more hardcore, more accurate, "more correct" – HotA will choose the simpler and prettier option, while Warlords chose the more hardcore option.
And this approach permeates specific decisions. If it's not possible to balance the game in a smart way, then here, have a Diamond – we'll add so much randomness that its quantity will hopefully reduce the randomness. And why, interestingly, do almost all (except maybe Lizardmen) top players, especially those who still remember Warlords, laugh at the Diamond (at best)? "Coincidence? I don't think so." The main direction influences all the specifics.
And I think that the main direction in both cases was chosen precisely because the founding fathers were more inclined towards that direction. That's why the formulation above about mathematics and taste turned out the way it did. (This is also an answer to Pikolan).