Scrapebox: How accurate is the harvester when it comes to reading your query?

I am wondering how accurate is it when it comes to reading your query?

I was testing things out and I put a long query with 'negative' keywords on it, example as follow:

site:.com "powered by wordpress" -"captcha' -"you must be logged in" -"comments closed" -"you must be signed in"

So I looked into my harvested URL and i was still able to find blogs with captcha, comments closed ones.

Am I doing something wrong here? Please advise

Any help with this?
Thanks

I am very sorry for the bump again, but desperately looking for an answer to this question. I have been testing it out for the entire evening and still couldnt figure it out.