A cool story from Sticky.

2StickyFingaz

The overall idea is correct. However, perhaps everything is a bit mixed up.
Faster and more dynamic gameplay is one thing, and development variability is another. These things are not directly related.

Development variability, as one of the components of tactical diversity, is in itself a necessary thing in any game. In fact, this was one of the main goals during the development of the game, along with balancing.
Changing the building tree in the Castles is one of the ways to move in this direction. At first glance, this looks attractive, but specific implementation examples need to be considered. Perhaps a single change to the tree will not be enough; it may also be necessary to change the bonuses.

Now about the game time. The advantages of fixes in this aspect are clear: the ability to balance with a smaller map size (this problem is also solved by a mirrored template), reducing the number of guards of neutral objects, and, consequently, the number of battles with these guards.
It should also be noted that the reduction in game time is limited not only by the problems of map balance and low variability with a small size, but also by the features of the hero's adventure gameplay. The fewer turns a game takes, the more significant the impact of the first turn on its outcome. From this point of view, it is logical to reduce the time of one turn (by reducing the number of battles with neutrals, and possibly reducing the number of heroes), rather than the total number of turns.

However, the topic of map selection can be developed. As an example, the PoC map is proposed because it has an element of randomness: random starting positions. That is, the idea is to combine the advantages of fixes and random maps. But there can be more random elements on the map. For example, random objects from a certain set in certain points on the map. From a technical point of view, this is quite feasible. It is also possible to give random properties to the objects themselves.

In practice, all this can look like a set of scripts that are enabled by a checkbox in the map options.

In general, first, it is necessary to develop specific options for the building tree scheme for at least a couple of Castles, script it, and test it. I can help with scripting if the options seem interesting to me.

In my opinion, making a new version for some illusory fix is nonsense. It is better to make a good map.

We are talking more about adjusting the gameplay for small maps, rather than for any specific fixes.