Is it worth checking the speed with a test in your city?

Question

Computer networks

Good day.
I argued a little with the technical support specialist of Rostelecom, he "crushed me with eloquence", but could not substantiate anything.
In short, he called a client with a problematic connection from Rostelecom, conducted a test, took measurements both on his car, and on him, and on another wasp (there was an error of wasps). All indicators are identical (somewhere around 27-30 Mbps).
Speedtest - 100-9, 2ip - 27-10
For the sake of interest, we turned on the torrent, the speed showed 4MB in a sect of 30-32Mbit, and that's what happened. The speed is less than advertised.
The dispute began with a different view of the connection. I tried to explain that calling your own communication center is stupid because the city is essentially a continuous local network (multi-level, but still) and the speed will always become identical to the port, so the best way to check is to connect to another city. Was he right about this?
The operator ridiculed 2ip as best he could, saying that he never shows the speed correctly (he just heard gygy from him). In my experience, I trust 2ip more than a test (always the right speed with a dozen operators (mts, star, yugratel, rostelecom, etc.)). Arriving home, I checked my real speed of 50-50 Mbps on 2ip, speedtest jumped up to 80 and remained at the level of 54 Mbps (for example, an error).

Answer 1

Both speedtest and 2ip do not show the real speed, but the weather on Mars - the result depends on too many reasons.
The speed according to the tariff is not provided to you by an abstract port of the provider switch, the maximum speed of which does not play any role for you, as a subscriber, but by a very real provider shaper, which is engaged in limiting the speed of access to the global network.
The best way to check the speed is to download a torrent with as many distributors as possible - 200-300 seeds.