It depends. I tend to be a bit more practical and assess everything on a case-by-case basis and weigh the risks, the benefits, and cost differentials and decide based on how that comes out.
For example, I have a Nest:
|Third parties know when I’m home
|Requires internet connectivity to remotely access and control
||Will still work locally without internet connectivity
|Someone could hack the Nest and muck with my climate control
||Google appears to take infosec seriously, the damage possible is minimal due to other mitigations
|Depends on a company maintaining their cloud services
||Nest is very popular giving Google incentive to keep their servers running
The cost differential was impossible to beat, my power company gave it to me for free. I’m OK with the mitigations above so I have a Nest.
NOTE: this is not an exhaustive analysis, I usually go into far more detail than this.
When choosing between options the fact that it doesn’t require a cloud service is one factor in favor of choosing one option over the other, but it is not necessarily the deciding factor.
I don’t know if I understand the question. If you mean whether avoiding cloud-based devices is a lost cause then I think the answer is yes, as time goes on you will have to compromise on capability and features and your choices in devices will become more and more limited.
It depends on what you perceve are the dangers from using wireless. From a security perspective, wireless can be less secure than wired in a lot of ways. They can be jammed. There are often vulnerabilities discovered in key aspects of the system that could let someone break into your system. There are some vulnerabilities in Zwave and more in Zigbee.
If your fears are health risks of wireless, IMHO, I think that might be a bit paranoid. Personally, I think based on my understanding of the studies published thus far and my understanding of the physics involved that the risks are quite low with wifi and the frequencies that most home automation wireless devices operate at. I’m no expert and I’m not going to try to convince you or anyone else. That is just the conclusion I’ve come to with what I know.
However, there are a lot of good reasons why one would choose wired over wireless. It is more reliable. It is easier to find and correct problems. It is faster. It is less susceptible to interference. From a security perspective it requires physical access to compromise. If it were practical, I’d run everything wired. But the cost benefit calculation doesn’t pan out for me. It would cost far too much for too little benefit to make it worth while.
Of course, wired is not a panacea for security. If you have a cheap Chinese IP camera that is exposed to the internet or downloading updates from a compromised update server it is just as exposed and vulnerabile as the one that does all this wirelessly.
At the end of the day I follow my training. I look at the risks (i.e. vulnerabilities * liklihood that it will be attacked * impact if the vulnerability is successfully attacked = risk) and see if there are any mitigations I can put in place and decide if the benefits outweigh the risks enough to accept the risk.