It's still 2FA. They're separate secrets.

At this point, it really depends on implementation, and the exploit.

It the exploit can get both in one go, I'd argue that it's technically 1FA. Else, no matter how trivial it to do both steps, it's 2FA. But then it pushes the question back to "what is a go at it"? A script? A remote file copy? Etc.

Kinda important technicality in my view, as separating them in the password management process is the first requirement to actually have two factors.

However, using two apps instead of two parts of the same app isn't much of an improvement. If the device is compromised, it doesn't matter much how many apps you split he data into. You can always use different passphrases, no biometrics, etc, but at that point, it's so inconvenient that you're just better off carrying two devices...

The point I'm doing here is: 2FA were originally supposed to be actually separated (other, offline device). However, for various reasons (cost, adoption, convenience, etc), apps were pushed instead. Now we have a regression where, in most cases, 2FA or MFA are often just a "single factor authentication with extra steps". As a matter of fact, true MFA was the main criteria when I selected my bank. And the day they force an app on me is the day I change banks.