[qasdfdsaq]

27 was at the time an instantaneous measure of your ability to fetch a certain type of data over the whole internet from a certain location. As been said several times, your local connection is one factor, the location you're connecting to is another, and the time of day also affects congestion.

The threads used by a speedtest site has nothing to do with threads as you think of them on a PC.

Since getting 16MB/sec is impossible on a 100mb connection (and the fact that we've already argued the fallacy of adding interdependant moving averages together) is a completely useless measure, and you can't base anything off it.

I'm really getting the feeling you don't fully understand how speed or speedtests work, so instead of presuming stuff as fact it'd be easier to explain to you if you just said what you don't understand.

From what I can tell, you got 27mb on a single-threaded London test at peak time, and a higher number that you don't specify on a flash based (probably multithreaded) test from Los Angeles. I don't get your point. All that says is your connection was faster off peak than it was at peak time... Why does that make either test wrong?

In fact, you seem to be claiming that because you did a flawed test and incorrectly calculated your 100mb connection could do 130mbps, a speedtest that says 27mbps is indefensible? I don't quite get your logic... Or are you saying that since you downloaded two files and got 16MB/sec combined, if you downloaded one file and got 8MB/sec, then that must be wrong? Cause I don't quite understand that either.