That all makes perfect sense I get it. I think though that you are not accounting for why they warned you. I think they did it specifically for the "Hi" messages and maybe sure it was for variations or repeated words but in that case make it clear what you are doing at the very least? If they allow bots like you mention on your other comment then you should be fine adding "welcome to hive" and since they do not seem to prosecute the bots then why should they treat you differently. The difference is the bots try to appear helpful?
Regardless I understand what you are saying in the broadscope but I think you need to just focus on the specifics which is hivewatchers, the platform and the rules being enforced.
In 2020, these past months, I have said hi around twenty times or just a few times a few different days this year. I reduced how often I wrote repeated comments in 2019 to escape bot detection. I sort of did this hi rampage recently in 2020, the past several weeks or so, to see if Hive had any bots yet like Steem did and I guess we do have @hivewatchers which might be a Hive version of Steem Cleaners or Cheetah or some of the others. I try to appear helpful too but good idea, pretend to be a welcome bot, I may try that and have done that at times. I vary in the different things I might say to people. Well, I'm not totally sure what triggered Hive Watchers to detect my messages, I don't totally know if it detects identical comments that are consecutive, that is duplicates, copies, one after the next, the same, or if this particular bot actually compares what you write with everything you have ever written in your life to look for patterns and go after you if you said thanks.
Saying Thanks
I sometimes reply with saying thanks. Just thanks. Well, I believe that some bots are not programmed to look at comment replies and the replies to the replies, and so on and so forth as it is nested in a thread like a forum. So, it might be that some of the bots are only looking at the comments that are replying directly to posts as opposed to scanning comments that are replying to other comments and comments of comments, etc.
Conclusion
I will probably be ok but I just wanted to let some people know the dangers of some bots, not all bots perhaps, but some bots can be dangerous sometimes.