hio77:
So figured i'd throw out some numbers on this end, having finally gained a spot of time to run the numbers accounting for stable connections this time!
...
final tibid of this analysis is, our lowest end of connections has seen no improvement, the top end already were pegging out and gained a couple of mbit through vectoring still to the top of the profile.
...
Perhaps an interesting comparison here would be "mean per speed band" in week before vectoring turned on vs mean for week starting say 2 weeks after vectoring turned on. Sort of like the plot below (except that uses your peak numbers) but for much narrower speed bands (say <10, 10-21, 20-30 ... 90-100, 100+). Again speed band based on speed before. Good to see both DL and UL curves plotted?
Plotted as above, the no improvement seen on your lowest end of connections (and at the top) does not show up
Overall, they do not look big changes, especially if only DL is seeing improvement :(





