Why wouldn't I use the code I posted? I'm using it and it works as expected. My code proxies external https to internal http, so all external traffic is secure. Yours is http to http and it does not do the same thing.
Also, as posted by the author of xbt_tracker not all torrent clients include the ip address in the query string, my code adds it to the proxied url just in case it isn't already there.
And, I have logged the announce also, it looks like this
Code:
1589673357 127.0.0.1 37032 GET /blahblahblah/announce?info_hash=blahblahblah&peer_id=-qB4030-xRH7k*70(3me&port=8999&uploaded=0&downloaded=0&left=0&corrupt=0&key=6E012D46&event=stopped&numwant=0&compact=1&no_peer_id=1&supportcrypto=1&redundant=0&ip=ipaddress HTTP/1.0
I grant you that the buffers may not be needed, but I did google a bit and thats what I found to be recommended.
I have no issue with saying your code works, but I do have an issue with you saying mine does not.
In my searches, neither method (proxy_pass, rewrite) is recommended over the other, so it comes down to what you prefer. I could not find any evidence that one is faster than the other.