proto-Indo-European geography.

From an Eastern or Central European perspective, some American (and even British) maps do look funny. For some unfathomable reason the Pripet Marshes, which actually cover a little more than 100 sq miles in the basin of the River Pripyat' (still admittedly the largest swamp in Europe), tend to loom larger than life on many of them, sometimes beginning just east of the Vistula and disappearing into the legendary lands beyond the right-hand border of the map of Europe.

Interesting. Yes, we do tend see the Pripet marshes as a vast vaguely defined swamp starting at the headwaters of the Vistula, extending to the headwaters of the Volga and Dneister. I imagine it's been greatly reduced in extent by modern drainage, but your comments are corrective. Nonetheless, it is a barrier.

[...]

Archaeological data show that neolithic cultures came to Poland via entry routes which were few and quite narrow: the Moravian Gate between the eastern end of the Sudeten and the West Carpathians, the route round the western end of the Sudeten (at the water gap of the Elbe, which separates the range from the Erzgebirge),

Yes. "The Moravian Gate" would be the gap created by the Morava River (of the north, the one that forms part of the boundary between Austria and Hungary, as well as Moravia and Bohemia). I have not heard of this term before.  I'm familiar with the Elbe valley approach. Both of these approaches are via Hungary or the Vienna Basin.
 

and the corridor along the Dniester and San valleys, by which the Vistula can be reached quite conveniently from the Pontic region.

This one I'm not really aware of. The books I've read make vague comments about going thru the Pripet Marshes, almost with a sense of wonder. This is the kind of stuff I'm looking for. No, we really do not learn Eastern European geography. This vagueness seems to infect just about every book on IE studies coming from West European and North American authors. I suspect we in the West would more appreciate the Eastern school's conclusions were we know the geographic realities.

[...]

The forests were rather dense, and if not literally impenetrable, they were at least difficult to traverse or to clear with the technologies then available.

Yes. The North European forest was a serious challenge. Even for later 'invaders'. The ancestors of Germanic and Balto-Slavic were left in relative isolation or were able to shrug off occasional instrusive linguistic elements.

They abounded in large animals like the aurochs and the European bison, red deer, elk (what you'd call moose), the wild woodland pony (closely related to the ancestor of the domestic horse), bears, wolfs, lynxes and wildboars. Human settlements clustered along the main rivers and were scattered over the large lakeland stretching between the lower Vistula and the Oder system.

My reading of Mallory suggests he is moving towards a 'disturbingly' late date for non-Anatolian PIE. If we are to have an undifferentiated common Indo-European speech continuum that agrees at all with the approximate usual dates for the emergence of the various daughter languages, you have to have the ancestral groups in place more or less by 2500 BCE, ready to spread out from the common center towards the Ural and Yenisey in the east to the headwaters of the Elbe in west, and regions south at a remarkably rapid pace.