Cat 5e = un-screened unshielded twisted pair. Good for 100Mhz and can run GigE Cat 6 = un-screened unshielded twisted pair. Good for 250Mhz (i think) and can do GigE Cat 6a = screened unshielded twisted pair. Doubles bandwidth but i think is still only good for GigE but be able to do 10GigE Cat 7 = screened shielded twisted pair. So thats means there is an overall screening as well as each pair has it's own screening. Good for somewhere around 700Mhz i think and can blast 10GigE down it.
Summary, it really doesn't have a place in many enviroment except for the data center or maybe tie cables from the patch panel to server rack or something like that.
PDL made a product called Lexcom for homes. It ran what is essentially Cat 7 cable just without the standard pair colourings. It can run audio, video, data and phone but in practice was rubbish and VERY expensive.
Both Cat 6 and Cat 6a can do 10 GbE (10GBaseT) but Cat 6 is limited to between 37-55 m runs whereas Cat 6a can do the full 100m. See for example the wikipedia article http://en.wikipedia.org/wiki/Category_6_cable . Even the shorter length may seem overkill for a home situation but since the depends on the cable quality and the standard of the installation etc, it's generally better to do Cat 6a if you actually want 10 GbE. (Given the current early state of copper 10 GbE I also don't think we really have enough info on real world performance of Cat 6 in 10 GbE networks anyway, particularly since as price and power usage comes down, performance may also degrade in borderline cases.) AFAIK, Cat 6a is generally recommended for new installations regardless of length where 10 GbE is the target.
I know some people would argue 10 GbE is overkill for home usage and if you hardly ever use LAN (particularly for transferring large files) then it probably is. But if you do use a LAN, then it's resonable to expect 10 GbE will take off on the desktop and home server front within a few years. People say the desktop or home server is going to die or that wifi is enough or you'll just never need 10 GbE but people were saying the same sort of stuff when 100Mb was the norm, before 1 GbE took off. And having being forced down to 100Mb due to a broken switch I can say you definitely do notice how slow it is compared to 1 GbE. There is also the possibility fibre will become the norm but then again both USB 3 and Thunderbolt (or whatever it's called now) were supposed to be using fibre and we can see that they still ended up with copper (at least initially in the case of Thunderbolt). Remember particularly with SSDs taking off you can resonably easily saturate a GbE link and even 3.5" HDDs are often able to.
Presuming you aren't planing to move within a few years, the only question is price. Since it's unlikely there will be a price crash, if the price difference for the cable + jacks etc is less then say 1.8 times it seems resonable to install Cat 6a. By the time 10 GbE is the norm the price of 6a would have dropped a fair bit but it's not going to be cheaper then Cat 6 and it's likely to be more since we're already quite a few years in to Cat 6 becoming common. So you're not likely to be saving money by upgrading in the future. Even double the price may be okay but one issue is the cable will likely degrade over time so it would probably be better to have a new Cat 6a installation in the future then one now if the total price ends up being the same. If you're having it professionally installed then perhaps even around 2-3 times the price is resonable since you're still going to have to pay the installer again.
(Note while it's resonable to expect Cat 6a is more likely to be able to handle beyond 10 GbE, such suggestions are IMO to hypothetical to be a big factor. Similarly Cat 7 (Class F) is way to hypothetical, particularly given the different requirements for proper termination.)
BTW, just wanted to mention that I wasn't trying to suggest you'd likely notice an advantage to 10 GbE if you can saturate a GbE link. If the difference is only 10% the difference would be too small to be significant. But if you can provide (including the computers etc of course and not just the raw speed of the HDs) even double of what a GbE link can provide 10 GbE (or dual bonded GbE) would likely be a noticable advantage. Good SSDs are already well there, HDs are still a while away but in a few years time when 10 GbE starts to make it in to ordinary desktop motherboards and computers and 10 GbE switches start to reach a resonable cost, they'll likely be there. Or to put it simply, there's no good reason to think 10 GbE isn't going to take off in the desktop/home server world just as GbE did (and as I said, there seemed to be many people saying similar things when fast ethernet was the norm).
In fact I mentioned clearly noticing 100Mb earlier, even my USB 2.0 external HD can do over double what a 100Mb link can provide, and we know how crap USB 2.0 is for mass transfers despite the hypothetical 480Mb.
And perhaps I should also clarify my comment on optical/fibre links. While it's possible fibre/optical may take off for device connections where you'll generally use a replaceable cable that can be made in bulk (or a throw away device) even though it hasn't yet despite repeated suggestions it would, the greater difficulties dealing with breaks and installing in the first place are likely to be a dampener on it taking off for SOHO and home ethernet connections (or really any in home cabling). Even if it's fine for you, most consumer devices aren't likely to use it for ethernet.