Are encapsulating Pools a threat to the Cardano-Network?

Based on the news that Sygnum Bank is now offering Cardano Staking and the positive message from Cardano Foundation CEO about it, I would like to raise the following question:

Is Cardano Staking in the true sense of the network possible if I do Staking through an exchange or a bank?

I think the answer is no.
As I understand it, if you hold your ADA on an exchange or bank, there is only a promise from the bank or exchange that you will be paid the equivalent of ADA at the prevailing rate if you want to sell your ADA, or that you can send ADA to your wallet or to any other address. In addition, exchanges offer lending, staking, and various derivatives. Whether the exchange or the bank buys ADA accordingly is not mandatory for now, right? And if it does, which seems to be the case at the moment, the person who “bought” the ADA on the exchange or at the bank is not the person who has an address in the Cardano network, but the exchange or the bank.
The bank has thus - and this explains the current structure of the staking pools with a majority of SPOs (67%), but less than 22% of the stakes - combined all ADA on its wallet or wallets and staked them in its own name.
If this is the case, it should also be the case that Binance, for example, can exercise voting rights for more than 11% of the stake, or am I mistaken?
If this is the case, this is a real threat to the integrity of the Cardano blockchain and all approaches to decentralization and governance, and strategies would have to be developed to counter this (if this is even possible due to the described voting rights).

If this is confirmed, which I actually think it will be, we have a real problem. I would even say, given the discussions on CIP-50, etc., that in terms of all the discussions around decentralization and strengthening the SPO, you don’t see the elephant in the room. There are financial vehicles besides classical exchanges, some of which nobody knows who is behind them, such as NEW GIRL, which act in a similar way, so the actual wallets have only 1 or 2 ADA positions and the structure of the ADA holders is not visible at all.

To give the whole thing a term, I would speak of encapsulation in this context, which manifests itself in encapsulating pools (e.g. Binance’s) or encapsulating wallets (the responsibility of the exchange or financial vehicle or bank). I think, the whole discussion about decentralisation and strengthening SPO does not differenciate enough between encapsulating MPO (like Binance, NEW GIRL) and non-encapsulating MPO (like ADALITE). It is a difficult question, of course, how to avoid this, but the parameter changes currently in CIP50, like pledge-leverage will not change the dominance of encapsulating pools, because it is very easy for them to split one ADA-Position in two positions, where one can be the pledge and the other used for flexible Position-Management from Epoch to Epoch resp. it would be also possible for them to manage some pools for Flexibility and the rest could have 1 position in pledge.

So, if I’m making a wrong assumption here, I’m happy to be proven wrong and would even be happy in the end. Or are there arguments on the part of the Cardano Network, which support such a Pool- resp. Wallet-Structure?