What about things beside spacehabs.
Do you think LEO constellations will expand in the next 5 years?
Yes, but we kind of already know exactly which ones and by how much. Mostly launched by SpaceX, I believe (Orbcomm and Iridium).
I mean Google is currently developing a satellite internet protocol.
Considering the advantages of such a set up (in LEO), satellite internet might actually be a viable alternative in some regions.
GSO internet is already used in several places. LEO broadband internet, while theoretically allowing both lower power, smaller antennae, and lower latency, is still kind of a ways off. Iridium's new constellation does have greater capability for handling data throughput, making higher bandwidth internet connections possible if you can afford it. (Iridium NEXT is supposed to launch on Falcon 9s from Vandenberg between 2015 and 2017.)
There are a few other companies like that out there, such as 03b (medium Earth orbit constellation of 8 satellites targeting the third world, or at least southern and low-latitude northern hemisphere) and "COMMstellaion" (LEO constellation of 78 microsats). But once they launch (if they get to the point of having enough money to launch), their launch demand won't be that high (they can last for a decade or two), and such a LEO or MEO constellation may well have a total mass that may actually be lower than a typical GSO constellation. They might cause an extra 10 launches every ten years or so for an average of 1 per year (multiple satellites per launch) for 03b and COMMstellation combined, but that's actually pretty small compared to the global commercial market.
There's continued growth there, but I wouldn't characterize it as explosive. We still have many of the constellations launched initially in the late 1990s, and they're only now starting to be really upgraded rather slowly.
To be honest, the demand isn't anywhere near what the annual upmass requirement of a well-stocked Bigelow-type station with regular visits would be (they've talked about a couple dozen flights per station per year, which may be what's required for it to be profitable).
One thing you have to keep in mind: there were NO commercial payload launches on US launch vehicles in 2011. None. All the commercial launch market is being handled by foreign companies. There are plenty of US commercial satellite manufacturers who would rather launch domestically if there was a competitively priced US launcher available, so if done right, there may well be room for a US-based commercial launch vehicle (ULA is too expensive, now that Delta II is retired). To support much more than one or two, there will have to be a large increase in launch demand somewhere.
(Yes, labor costs are higher in the US, but there are other advantages. The cost difference has been lowered in the last few years, as wages in the rest of the world have increased. Increased automation and different manufacturing techniques can partially make up for the difference in labor costs.)