Good board games can be translated into good games relatively easily because mechanics can be coded in and adapted in a very open-ended way. The inverse however is demonstrably untrue, as most video games would make for terrible board games. The problem is neither balance nor game design, it's the simple fact that video games are structurally complicated and rely on an enormous amount of automation and simultaneity of events/mechanics to function, something that a real board game would need actual humans to do. To put it simply, even the simplest strategy video game could become one of the most hardcore board games if translated with complete accuracy. This is why board game adaptations of video games (incidentally, also the Civilization ones) are grossly simplified and smaller-scale versions of the originals.
The idea that the snowball effect is a worrisome effect of Civ VI design is also unsubstantiated imo, even when taking board games into account. The vast majority of highly successful strategy board games are extremely snowbally: Catan, Risk, Twilight Imperium etc. The problem with Civ VI's late game is more of a matter of engagement and management: There is great fun and challenge to be had in administering a winning situation and it comes with its own unique set of obstacles and difficulties. In games where rubber-band mechanics reign supreme, the comeback of your foes is the main threat; in snowbally games it's your own possible blunders and mismanagement. Civ VI is simply too easy in the late game, making management non-risky and formulaic.
If Firaxis changes the dynamics of the late game to make it actually stressful to manage your empire and for internal factors to pose real threats to your victory even as your foes seem woefully incompetent at catching up, then the game will improve drastically and people will actually be willing to settle with the sub-par uncompetitive AI.
. If AI knows how to 1) take city immediately after nuking, or 2) nuke multiple targets, it would be a lot more threatening.
