The more I think about this feature, the more I am questioning it.
(...)
I don't see the realism factor nor do I see any vital gameplay factor.
I agree, but the thing with the (growing) improvements is about the same way as with all the other features:
It
is truly
difficult to have a clever balance between history (accuracy) and good gameplay (fun, strategies, etc.).
You are right CB, that also the (vanilla) main principle of having "splitted" work factors (improvements on tiles, workers/experts in city-screens) doesn't make it easier to have realism in, but as this principle is the core of Col-Mechanics (and I however like it...) I wouldn't mind nor change...
automatically growing (advanced) improvements:
The only factor now is IMHO to protect your advanded improvement, once they got developed (similar to Civ4, when you didn't want your enemies to plunder your developed cottage/town-improvements).
And the mechanics of development
could be reasonable because after being worked by several generations, the farms/plantations
got actually better as there is a saving (and growing) of Know-How
from the fathers to the sons.
In general, a farmer of 1500 wasn't less efficient than a farmer of 1700. A field being plowed in 1500 didn't give less yield than the same field at 1700.
In the game, now it does.
Sorry, but from the point of agricultural history, this may not be
strictly correct overall

For sure, the main and really big steps in agricultural development didn't happen within the timeline of Col (use of wheeled plough: Roman era, crop rotation: middle ages, chemical fertilizer: 19th Century) but neverless there was
some considerable development during that period:
1. By having new crops from America (like potatoes) farmers began to improve the three-field-rotation and hereby developed the four-field-rotation.
2. By improving the wheeled plough during the 18th Century, farmers got better tools to work with, what resulted in more benefit/harvest
So, I am not against a second level of improvement in general (as they
are reasonable from point of history), but I agree with CB that the way how they are constructed/implemented
may need an overhaul. Maybe there are better solutions than the current "automatic development"...?
-> Should the 2nd level improvements perhaps be buildable by a special unit (kind of Engineer)?
-> Should constructing of 2nd level improvements be expensive and rare?
-> Any good ideas?
