I'll agree to disagree. For policies trad and rationality are definitely stronger than the others, but there aren't really any "useless" policies. Even piety is useful if you want to pursue a religious victory. Other than filling out trad and rationality the other two or three or four (Poland) trees you take are up to your discretion.
Also not really sure what you mean with buildings. Some are universally good while others are more geared to certain playstyles (like aqueducts for tall cities), but I can't think of a building choice in BNW that is useless.
Beliefs yes. There's not much point in founding a religion after someone else takes Tithing.
I guess what I mean to say is that while Civ V's systems are simple, the core mechanics work. Which is good. The main problem is that Firaxis didn't do balance very well.
Policies: Saying that one early game tree is essentially mandatory to fill, therefore meaning that your first 6 policies are basically locked in from the start, doesn't seem like a very good state of the game in my mind. Liberty was fine, but the unhappiness penalties were hindering it too much. Honor was... ugh, hit or miss, as it depended on how many barbarians you had nearby. It also didn't help your infrastructure at all. Piety didn't have a way to generate culture until the tree was complete, so it had a very weird gameplan where you needed a culture pantheon but relied on getting a religion.
Buildings: Science buildings are definitely the core. The game literally revolved around stacking the % bonuses of science buildings.
Leaders: I felt that in vanilla, some leaders were more or less unplayable due to just how dull they were. We obviously knew Korea and Babylon were the best, since science is king in vanilla. But Byzantine literally didn't get an ability if the AI got every religion (which can easily happen since there are basically 3 Pantheons that are faith producing). The Dutch were more or less begging to get a Marsh/Floodplains for a Polder, since their UA was... questionable. Indonesia worked on very specific maps only, and their unique unit could actually be
weaker than a regular one, making it a seriously weird experience. The Iroquous UA sometimes failed to work and their UB was
worse than the generic building due to a lack of a % bonus you relied on.
Gameplay: Higher difficulties were made by catapulting the AI forward from the start and playing 'try to catch up and not get overrun'. In VP, the AI starts with much lower bonuses, but catches up to the player instead. Spawning near someone like Attila means you had about 40 turns to prepare for the inevitable war. Which the AI sucked at, bust quantity can sometimes beat quality.
The core mechanics worked, sure, but the gameplay didn't differ between one game and the other enough to matter. It mostly boiled down to turtling the first 100-150 turns until you caught up and got ahead of the AI, then rushed every % science boost you had. Playing wide was challenging at best, infuriating at worst (Oh, I lost a point of happiness and now none of my cities work, cool). You either filled Tradition, or you tried to cheese Honor on the Aztecs (which was fun, but ineffective).
In the end, VP feels like a more complete and enjoyable game with more varied gamestyles and choices due to how those gamestyles have meaningful decisions and are not completely overshadowed by the shallow mechanics.