I'm sure you're right on the PC being
a factor. But the above results to the same server from the same high spec don't support your the specific scope of your conclusion.
BTW, I'm using BT Infinity today (not that it matters to this discussion). The other factors are the pipes at the Speedtest server, the number of users converging on those pipes, activity on your local node from minute to minure and so on. Speedtest.net is nowadays a bigger scarer than that wretched Pingtest.net and 10 dBmv downstream power!