I think it may be a subjective response.

First of all, many BIPs don't change consensus rules, many don't even change how your node runs. A great example of this might be BIP 39 which is how we got mnenomic seed phrases rather than long strings of base58 characters. No one is forced to use BIP 39 seed words -- some wallets still don't (notably Bitkey and Muun). But I think it's fair to say BIP 39 has led to some useful things in Bitcoin.

In the above list, a great example of this is the BIP proposal for Standard Encrypted Wallet Payloads. This BIP suggests a scheme for encrypting your wallet descriptor and labels with your private keys so you can safely store this important information in less secure environments and so you don't have to keep track of two pieces of info in order to recover a wallet safely.

Does this BIP change Bitcoin? I suppose in the sense that many wallets will likely adopt it. But I see this as a pretty good thing. Improved usability that doesn't require consensus change sounds great to me.

Another example of a BIP like this is Craig Raw's BIP 329 which proposes a standard for wallet label data. This is an excellent change to how Bitcoin wallets work. I don't see it as a threat in any way.

Other BIPs do change Bitcoin consensus. An example in the above list is Rusty Russel's Great Script Restoration proposal. In this case though many of the changes are actually undoing changes that other people (Satoshi) did earlier in Bitcoin's history. Do you consider changes Satoshi made as "not changing Bitcoin"?

Finally, I'd note that there are currently almost 200 numbered BIPs in the Bitcoin repo. This doesn't mean they are all accepted or activated or widely in use. But at least 70 of these BIPs are in their Final or Active state, which means they have seen fairly wide adoption.

As far as I can tell, Bitcoin is doing what it has always done which is continue to change as more and more people come to use it.