From what I can see, the primary driver for this is actually plumbing to 802.11AC Wave2 and future APs.
With wave 2, there's finally a real chance a 1Gb/s link to the AP will be insufficient. Plus, Cisco and a few others are offering APs now that can simultaneously run 2 independent 5 Ghz radios in the same AP to give better density in a physical space without lots of hardware. Ans, they need something that can be deployed without re-running cable..
Most of the guys who do this professionally understands that a modest jump in speed is only saving you for one round of upgrades at best. The compelling thing of course is POE, if it wasn't for the issue of how to power the little bastards, running fiber for AP's would be the logical reasonably future-proof fix. We're currently capped at 1
While I don't see 10GbaseT ever making it huge in the datacenter, with 10/25/50/100 QSFP 28 switches and cards coming on the market, I think you're going to have a lot more mainstream hardware options if you stick with 1G or 10G. Particularly since, as others have said, the NBase standards aren't ratified yet. Only Cisco has announced prestandard switches so far.
It will be interesting to see how the prosumer market adopts NBase and/or continues to adopt 10GbaseT. So far it seems like most of the NBase switch support is derated 10G chips, with a custom PHY that does 10,1 and 100m plus the NBase speeds.
We've already seen this. It isn't a big hit. Netgear bet big on 10Gbase-T back in 2013 with their "low cost" options, and as far as I can tell, response was tepid.
This gets back to a meta-issue, which is, how much bandwidth is actually "sufficient". I've talked before how we rolled through 10Mbps (standard in 1993), 100Mbps (1996), 1Gbps (1999), and then 10Gbps (2002) ethernet on a fairly consistent 3 year cycle, with commodity hardware availability trailing by a few years in each case, EXCEPT for 10Gbps. What seems to have happened is we finally got to a networking speed that was "fast enough" for most purposes, because if that wasn't true, then we'd have seen 10Gbase-T get mainstream years ago.
We don't seem to have a big call to have desktops connected at >1Gbps. It logically follows that there isn't a big call to have laptops do so either, so the primary difference between those situations from a cabling point of view seems to be that an AP is serving multiple clients. Which is what you said in the first place. But really, wifi is typically serving a large number of lowish bandwidth devices, and usually with that you just don't have a good chance of peaking out a wireless network. That might be different if you've got one high-performance client that has an AP to itself, but that brings us back to the "desktops are fine at 1Gbps" issue.
I suspect that the average WLAN is going to find that their client-heavy Wave 2 networks aren't going to find them backhauling > 1Gbps on a constant basis, and that even high-performance clients on a 1:1 basis are quite possibly okay at 1Gbps.
This opinion seems to be shared by others:
http://www.revolutionwifi.net/revolutionwifi/?offset=1422633299171
Since the big driver here is POE, and there isn't yet a 10Gbase-T POE standard, yes, that's trouble. But it is also worth noting that the whole Nbase thing itself might quite possibly turn into another wireless-N debacle:
http://blog.planetechusa.com/2014/12/what-the-nbase-t-and-mgbase-t-alliances-dont-want-you-to-know/
Even shares the letter N.
I think the smart money is on avoiding Nbase unless there's a compelling need and you feel completely comfortable with trashing the stuff in a year or two. There will inevitably be some folks for whom the product fills an actual need, but most of what I'm seeing is vague fearmongering about how 1Gbps "isn't enough" in what appears to be a blatant attempt to upsell existing sites new technology that they don't actually need and won't actually benefit from.
