Hi everyone,

i am developing an isometric game with the Unity graphics engine, and being the first time I approach the programming of a video game I have some perplexities, maybe even a bit stupid, but I would not like to do a bad job.
I was able to recreate a hybrid tilemap / sprites environment via prefabs, only I was unable to create a full map with the tilemap system for several reasons:
The first is because I need a dynamic scenario in which some objects could move.
The second, the most important, is due to the positioning of the walls with the "snapping grid", they degrade along the X axis and increase along the Y axis creating a "step" effect (the asset is oriented at 35 °).

Later I discovered that the same single tilepalletes can be positioned inside the prefabs and unlike the brush, they move freely, but with them the entire grid also moves.
It is a correct method to put a layer for each tilepallettes, but above all, there is convenience in terms of memory, to build several layers of tilepallettes in a single prefab, or for the objects and walls, I continue to simply use the sprites, without have to worry about memory consumption?
I would like your opinion, thank you.

Your graphical problem isn't obvious: the screenshot shows a perfect, properly antialiased, 35° wall.
There is only an unfortunate choice of floor texture (in real houses, tiles are aligned with walls) that can be corrected with mere asset adjustments (skew the current image 35°, wrap around the corners to get a rectangle again, and align it properly).