No. The amount of centralization of non-mining validators is completely orthogonal to the amount of centralization of miners. The two metrics have nothing to do with each other. And to repeat (I sound like a broken record, but you keep missing this reiterated point), NON-MINING VALIDATORS HAVE NO POWER ON THE BITCOIN NETWORK.
As stated above, No. Not yes. As for your 'then': Merchants are likely to themselves benefit from running non-mining validators. This allows them to transact trustlessly, and to monitor for double-spends. Users? Meh. Either way, NON-MINING VALIDATORS HAVE NO POWER ON THE BITCOIN NETWORK.
Again, no. As for your dependent clause, no again. It is not worse, but it is also not better. For the only entities that have any power upon the Bitcoin network -- i.e., the miners -- are almost fully interconnected, and therefore have no need for non-mining entities to relay anything on their behalf.
Ok but if non-mining full nodes have no power on the network then what was the UASF and the NO2X movements? Were they not comprised of users running non-mining full nodes?