I think your points are totally valid. It’s certainly something that can be over used and/or misused. In a few case this approach will likely improve UX for example on Google maps where the user can scroll content underneath and maximum space is required, so no nav bars.
This approach is not a design pattern, it’s not to be use throughout the app, just in specific places where’s validated. Test it, you can’t just follow trends thinking — “Google do it, it must be good UX”. As my latest blog post mentions (1st para), you can’t just take inspiration from something without thinking and validating it first.
I feel the argument about the FABs blocking content is one that could be easily mitigated by allowing it to be dragged. A bit like Facebook’s messages app’s ‘coin’ avatars. I found them irritating being on my home screen though. But whether user know to drag them is something for testing.
So the bottom line is:
- FABs can be useful in limited situations, when they are they should be tested and validated, with research to ensure they are a key task.
- They’re not a design pattern, don’t apply this to your whole site / app.
- Consider use carefully and allow users to move the buttons where they obscure content.