Modern washers have a timeout (in case the water-fill sensor fails) which is around 10 minutes, but you'll have to check the manufacturer to get the exact timeout for your washer as well as how much water it needs to fill at your preferred load settings.
If (DesiredCapacity / HeaterGPM) < TimeoutInMinutes
...then you'll be ok (I wouldn't cut it very close though - it would be a bummer to see it time out one millimeter short of full) So if you had a 10min timeout and needed 10gal to do a tiny load with a 1.5GPM heater:
10gal / 1.5gpm < 10min ? True.
... to be fair, a maximum load of a washer is typically 40gal so the concerns would be reasonable.
Because of all of these variables, those who make the washer would know best since they would know these variables.
Front Loaders spin at an incredibly high rate compared to top loads. Because of this, the amount of vibration that is transferred to the structure is immense compared to the top loaders of days past.
A concrete slab is the ideal substrate, as it will absorb all vibration. But if the home has a solid joist system and subfloor, they are viable. However, there will still be noticeable vibration. We have a 1989 built home (also in Puget Sound) that, alas, doesn't handle the front-loader washer we just bought a few months ago as well as we'd like. I will be installing lolly columns in the crawl space to reinforce the laundry room floor. If that doesn't cure things, it looks like I'll eventually be building a laundry room in our garage.
The reason they are not allowed on second floors has nothing to do with leaking (as that's a risk of ANY washer, of course) but because the live load requirements for residential construction differ between the main floor and the second floor.
I can't find a direct source to cite (if anyone can find one, please share!) but IIRC, the live load requirements for the main floor of a house is 40psf, but the live load requirements for sleeping areas on a second floor are less at 30psf.
A front load washing machine, in spin cycle, is a very "live load"*. :)
The bottom line is that in the US, the typical construction of second floors aren't equipped to handle the abuse of a front-loader washing machine.
You could probably get an exception if you reinforce the flooring in the second floor room that will house the washer. But that's likely not going to be cheap--ultimately costing you more than the savings in energy you get with the front loader. I believe PSE's top loader option is still a high efficiency appliance, though, so you'll still see a savings in energy with that. It just won't rattle your house quite as much.
UPDATE:
Found the load tables. Here's an excerpt:
From the International Residential Building Codes
TABLE R301.5 MINIMUM UNIFORMLY
DISTRIBUTED LIVE LOADS
(in pounds per square foot)
USE LIVE LOAD
- - - - - - - - - - - - - - - - - - - - - - -
Habitable attics and attics
served with fixed stairs 30
Rooms other than sleeping room 40
Sleeping rooms 30
(*) NOTE: As ben points out--and as I attempted to point out with the smiley face--the fact that washer vibrates a lot isn't what makes it a 'live load' necessarily in the eyes of building code. My point was that the upper floor is just built to a different standard than the main floor. That said, I do see the broad definition of live load also includes vibration: http://en.wikipedia.org/wiki/Structural_load#Live_loads
Best Answer
My Sears Kenmore top loader washer machine tub has a construction where there is a donut shaped compartment built around the bottom of the tub. It has water inside of it!!
A couple of years ago when I replaced the tub bearings I was mildly surprised to note this interesting construction. It did not leak or pour out when I had removed the tub during the repair. I can only guess that the water plays a role in the dynamics of the tub motion in conjunction with baffles inside the water compartment.
So I would guess that this is normal for your GE washer tub as well.