The statistics are interesting
Let us say, as a working hypothesis that, roughly speaking, Win 10 traffic doubles at the weekends and indicates a market penetration of the retail sector of somewhere between 25 and 30% in the US market as indicated by the graph of the US government's figures. If we then also accept (again, roughly speaking, as a hypothesis) that the ballpark figures provided by Netmarketshare and StatCounter are a reasonable indication of reality as far as the overall average for the combined retail and enterprise markets then the implication is that Win 10's penetration of the enterprise market at present is approximately half that of the overall average. With Netmarketshare's figure for February at at just under 13% and StatCounter's figure at just under 15% this suggests that the degree of penetration of the enterprise market is in the range 6 - 8 %-age points (again, in the US market). These figures (if they reflect reality) would seem to suggest that at this stage in the release cycle of a new Windows interation the figures are, from Redmond's point of view, respectable (given that enterprise always adopts at a much slower rate that private customers and we are only 6 months or so into this release cycle). However, the coming 12 months will be key. If we do not see an enterprise share of about 15 - 20 %-age points in a year's time then Redmond has a problem.