The law in its current form was passed in 2012 and was specifically restricted to devices that have the purpose of spying. We can safely assume that the lawmakers were aware of current technology.
Certainly not everybody. This law was last changed as a part of a large telecommunication bill. The proceedings show that they are aware that technology can move very fast (thus the law needs to be written as open as possible) and it's hard to precisely differentiate between e.g. phones (or new form of legitimate technology) and dedicated spying devices (thus the law needs to be written as specific as possible). Given those goals I think they made a pretty good job.
You're saying nobody in Germany uses "ok google" or "hey siri" hotwords? What constitutes an "active microphone"? Is it only when audio data is being recorded or streamed, or is it the mere act of processing audio with any type of software? What if the software is malicious, running in the background and not indicating it's actively recording/streaming/processing all the time?
I'm asking this mostly rhetorically, but just trying to point out that when most things these day are controlled by software that is remotely and automatically updated and installed (including firmware, baseband software, and background apps), it's nearly impossible to say something like "yes, this device has a microphone, but it's not 'listening' right now", let alone define it legally.
At the same time, nobody with a smartphone actually knows whether or not they're carrying an active microphone around or who could be listening to what's being recorded.
The purpose of the "Hey Siri" and "Okay, Google" functionality is to listen to those specific phrases. It is not intended to be used to covertly record people and can not be used by a regular consumer to do so. The law does not concern itself with technicalities such as what constitutes an "active microphone". What matters is intent. That's a pretty easy question to answer for those watches. This functionality is literally a bullet-point in their ads.
And yet we've seen similar technology in other products, like Samsung TVs, being used to gather huge amounts of information for the U.S. government. From a security standpoint, understanding the capabilities is really all that's relevant. If it's possible that it could be used in that way, then it probably can't be trusted not to be used in that way. Especially since these devices receive OTA updates and there are a number of ways to attack them on a relatively large scale by doing things like spoofing a cell tower.
If the concern is individuals using their own devices intentionally for spying purposes...that's relatively easy for a non-technical person to do with a smartphone if they want to.