Assuming you're not running major data service out of your house, what's the point of diminishing return for connectivity?
That would depend on the price, wouldn't it? If the marginal cost of 10 Gbps vs 1 Gbps is negligible, by all means, provide 10 Gbps. 10 Gbps ethernet over copper (for use within the residence to be able to take advantage of this speed) is still at the margins in terms of price, but that's mostly for the same reasons that 1 Gbps was so expensive for so long. If only "enterprise" uses it, it stays expensive, because business, as always, charges all the traffic will bear, and business customers like to pay more because they think that means they're getting something valuable.
Once residences started using 1 Gbps, the price dropped and dropped and dropped and now you can get a very good quality 24 port 1 Gbps ethernet switch for less than $100. 10 Gbps will follow the same trajectory, but the demand has to be there. This is the first move towards creating that demand.
Other people have pointed out that the server side won't talk to you at 10 Gbps anyway. You're throttled by the server at far lower than that. I've pointed it out myself for the past few years. But we know that the backbone bandwidth is in the ground, unlit, to support far higher outbound throughput from data centers. There's just no demand, and it saves on server hardware. Again, this is a move towards creating demand.
Somebody has to be first, and it has to be on the demand side. This is one of the first, at least in the US.