I think the best way to protect from this is two-fold. 1) Cybersecurity, and specifically the art of finding bugs, needs to be made an essential part of Bitcoin education. 2) The community needs to start producing its own chips. Whether by 3D printing or some other way.
Open source chips would be a good thing, but I think at the current state of things we are unlikely to see devices with such any time soon. The problem primarily lies in the trade off between usability and security. One example that I can think of is that multisignature wallets aren't used nearly enough. Multisig must be most standardized, with cross-device compatibility and good UI. Having security that is obscure and hard to use by the average person is pointless.
In that sense, you can imagine a 2 out of 3 multi-signature setup that uses 3 different hardware devices by 3 different manufacturers who use 3 different chips. It is extremely unlikely that an hack would happen in this scenario, especially not one that was not exposed before a second hardware device is compromised. Why? Given how hard this is, a successful compromise of 1 brand would likely be executed before a second one would occur. Clearly the usability of such a setup is terrible compared to something fast as a browser extension. However, for long-term cold storage does it really matter if you need 5 minutes to create a transaction knowing how much security you have?
When it comes to your small or every day wallet, I think you should use it with the expectation that you can be hacked because there are so many ways for this to happen. Unless of course, your daily wallet is also a hardware wallet. Then that is a different story.