Auto-translated
Thank you for your feedback, AmberSoler!
I will take your opinion into account.
However, you should understand that the aspect you suggested, I had in mind in the section about heroes (in your example - innate hero spells) and maps (I meant - the landscape). What you provided as an example, in my opinion, can easily fit into the existing framework (a global conclusion implies this). This is obvious to me (if you had carefully studied the post, it would be obvious to you too).
The question that arises from you (given the conclusions you have drawn) is - why did I do this?
Read the posts comparing cities...
Regarding a specific map for comparison - this is an ideal case. Specificity eliminates many things in the comparison and adds nuances, taking into account the data already obtained according to the template (or do you have a Gem without troops, etc.?).
It is interesting what you think I mean when I say that the conclusion will be ambiguous.
Well, finally, I will say that you do not fully understand the meaning and practicality of this template. The practicality here lies in weighing all the pros and cons. Assessing the advantages and disadvantages of castles (without missing anything). And if you wish, taking into account specific features (maps, heroes, etc.), apply the template to specifics.
For clarity, I will give an example with your Physmig. Let's take the section - analysis of the effectiveness of units of each castle and the overall impression. It is obvious to me that the effectiveness of a particular unit can be highly dependent on specifics. And your conclusion that "at first glance, the human army will easily defeat an equally powerful necromancer hero" (presumably referring to the army of the necromancer hero) in relation to specific map features, where equally powerful troops cannot exist in principle, looks "impractical," as does the entire section where the effectiveness of units in general is considered. But the whole point is that this information is valuable in itself, in general (your overall impression), regardless of a specific case. And already using the information obtained, everyone can apply it in a specific case.
P.S. I had the opportunity to thank you for Physmig. Thank you again for your feedback.
Sincerely