I've seen a few posts in this thread (and others) pointing out that 2FA can only do 100% of its job if it's implemented correctly (i.e. uncorrelated factors on fully separate devices). That's true, but I think it's worth pointing out that there are still important security benefits, even when it's done "wrong" (i.e. both factors, like your password manager and authenticator app, on a single device).
Like witcher_sense said above, just because you can compromise a device in one way, it doesn't necessarily follow that you can compromise it in other ways. Certain kinds of malware may be able to get to your password (keyloggers, clipboard sniffers, etc.) without ever being able to compromise your second factor, even if it's on the same device.
Also, some attack vectors don't rely on compromising your device at all (like phishing), and in those cases 2FA is a game changer, even in its "bad" form.
Full-strength 2FA is laudable and should be the goal, but even weak 2FA is beneficial and worth having.
I'd say that weak 2FA is only acceptable if A) you know the risks, and accept them B) It's a temporary measure until you can implement better security. I'm a firm believer that security shouldn't be compromised on unless the drawbacks are much, much higher than the benefits. For example, security should only be compromised for convenience, however if you compromise too much, you no longer have security. If you don't have enough convenience it's likely to cause you problems, i.e having a key file stored 100 miles away from you, in a field is probably going to cause more harm than good if you need that to access funds.