To me, one thing is sure for me. It works. Proved many many times for the last 10 years.
Now I am not really interested into running such tests mostly because I am not motivated enough (I am a programmer) and also because for myself I don't need any further proof for my own use.
For someone that's scientifically minded anecdotes like the ones you've described would actually count against the idea, not for it.
Why? Because human beings are very bad at intuitively interpreting information methodologically, this is why we have things like double-blind trials in the first place. It's very easy to subconsciously stack the odds in your own favor.
I'd encourage you to read up on James Randi and some of the people that compete for his million dollar challenge. Everyone that takes it is equally convinced (as you are) of their woo-woo ideas, and all of them are proven equally wrong.
You aren't immune from deluding yourself, and neither am I. Nobody is, that's why we have a process to factor human error out of the equation. It's called science.
In this case we aren't talking about some harmless and quaint idea. If you continue foistering homeopathy on your loved ones, there's a real possibility that the magical thinking will rub off on on them. And they'll subsequently neglect to seek real medical attention when they have a problem that can't be solved by a glass of water and a friendly smile.
There's a fantastic video online somewhere of a martial arts instructor that had built a practice out of some form of "energetic combat", where he would run around and wave his arms, and his students -- somehow convinced that it worked -- would helpfully fall over.
Then one day there was a challenge from a more traditional martial practitioner. The sensei was quite surprised when his unconvinced opponent did not fall down like he was supposed to.
And even more surprised when he got smacked in the face.
Unfortunately, rather than grapple with the realization that he had been practicing nonsense for years, IIRC he came up with some woo-woo excuse for why it didn't work that day, or on that opponent.
Now I am not really interested into running such tests mostly because I am not motivated enough (I am a programmer) and also because for myself I don't need any further proof for my own use.