Yes it is really difficult to decide what to draw with all the changes we've made to the environment.
The state of archaeology in the region also doesn't help. But in general we're starting to develop a better appreciation for the impressive irrigation systems present in this region during the ancient past, as well as the interactions between nomads, pastoralists, and the settled peoples. But yeah, especially the mind-bogglingly stupid Soviet period has done so much harm to the hydrology of the region.
I also assumed modern agricultural land would have been a sort of floodplain back then, though I don't know if that is true.
You're correct in that assumption, though the extent of some oasis/floodplain regions was different. Chorasmia especially had a massive (but super fragile) irrigation system that was even more extensive then it is today. Similarly, the delta of the lower Jaxartes seems to have been of a similar size, though lacking the irrigation networks and thus being more swampy, since it was primarily the territory of the Massagetae. The Ferghana valley and Chach region too were very fertile, though here the Nomad-Settled interplay seemed similar to that in Chorasmia where nomads and settled peoples were highly reliant on one another.
Though the big farmland region between Myrzakent and Jizak is a modern version - this region used to be the ''Hungry Steppe''. In general, any lakes are a modern appearance (outside of the Aral Sea, Caspian Sea, and lake Sariqamish which all are much smaller then they were in Imperator's time).
If you have more suggestions I am all ears.
The Farmland in Samti (ID: 6759) still seems to be there, and which has 0 historical basis (
This is the area that province covers, roughly). I'd further widen the Oxus delta, which seems to have had about half a dozen major channels going into both the Aral Sea and lake Sariqamish, with many smaller channels rejoining (still not possible I believe, sadly) or ending in swamps. The right bank of the Oxus
(inbetween Hazorasp and the Sultan Uveys mountains) seems to have been much more extensive as well during this time (being practically annihilated following the Mongol conquest, and only being reused for agriculture in modern times). Though the area past the eastern end of the mountain probably was little used and swampy all the way to the Aral. (Note that work on the east bank has been steadily revealing more and more impressive evidence of ancient Chorsmia, including a hypostele hall to rival those in Persis, but much more is left to be uncovered.)
In terms of the northern Oxus marshlands (with some lightly wooded swamps interspersed), roughly north of the modern Nukus is where these started, with the area south of that being mostly farmland. Roughly
this region, though the farmland and irrigation networks extended further west to Sariqamish into what is now desert. With the edges of the Ust-Yurt plateau being bordered by marshes. The Oxus past the Turkmenabat/Farab stretch and Chorasmia seems to have been quite barren, with a few marshy woodlands and mostly villages and nomads living there.
As for the Zeravshan/Polytimetus/Sughd, the lower (Bukharra, ancient Numjiket) oasis seems to have extended further west then it does now, and end in the alluvian fan at Paykend. I cannot find my notes right now on the state of the canal systems in the upper (Samarqand) area, but I recall the canal systems there were still being developed and extended massively during antiquity, though the main Dargom section was already ancient by this time.
As for the Ferghana and Jaxartes, still working on this region myself. Though the area around Chach (ingame Maidanchach,
this area) had high argicultural potential, though this was only realised much later since unlike the Ferghana valley this region remained mostly pastural. In terms of the lower Jaxartes, the branch that goes south (and then west into the Aral) ends in the desert in game right now, but it is this region where we have the most evidence of settled life along the Jaxartes in this period.
Finally, I'd add some more routes through the Qarakum, Qizilqum, historically used by the various nomad groups for quick raids into the settled regions. And get rid of the Wusun, Sogdia, and Tayuan tags and take away all Bactrian provinces past Alexandria Eschate and on the north side of the Jaxartes. The Ferghana valley and upper Jaxartes were under Amyrgian control, and the long valley controlled by Sogdia and Wusun is where the Sakia (Properly named Sacae) tag belongs. It is from there they were pushed out infront of the Yuezhi (Tocharian) migration which eventually saw them settle in modern Sistan (Sakastana) and found the Indo-Scythian Kingdom. The stretch of land around Shymkent was possibly Sacaraucae land, with various other groups probably part of the Massagetae culture along the rest of te river and on the Aral shore.
There are many more suggestions for this region I could make, but my time is up (dinner is served

) and in general just suggest you keep doing what you are doing, and occasionally check an atlas or even just Google Maps when it comes to
any mountainous areas, since as you have noticed they tend to be one big mess in Imperator.
BTW how come you know so much about ancient hydrology in that region of the world?
I have a big interest in historical geography. And I am reading up a lot on Central Asia right now since I will be tackling that region soon myself, given how uncharacteristic it is. Though I am currently still focussed on the southern Caucasus, Armenian Highlands and Anatolia.