ABC's two-screen My Generation app for iPad, released in September.

Not so long ago, it was pretty hard to watch online streaming video on anything but a personal computer. It was a big deal that Apple’s iPhone came packed with an integrated app for YouTube. Last year, the walls came down, as video services and device makers rolled out new native applications for one machine after another, from phones and tablets to smart TVs and settop boxes.

I love this kind of gadget news. I lived off it writing for Gadget Lab last fall. But shiny apps and feature wars are one thing — whether viewers actually use these services and how they interact with them is very different. Bit by bit, the data is starting to come in.

Video gobbles up bandwidth — it’s just a lot of information, especially at higher resolutions — so it’s no surprise that bandwidth studies now show video topping all other uses. What may be surprising, though, is the concentration of that bandwidth use in just a few companies’ services.

Just as Netflix owns broadband, YouTube owns mobile. A new Allot Communications report credits YouTube with consuming 22% of all mobile bandwidth globally, and 52% of all mobile streaming video. Besides its deep integration in Android and iOS, YouTube’s videos are perfect for mobile devices’ typically smaller screens and and users’ shorter snatches of time available to watch.

As for long-form video like TV shows or movies, Nielsen’s online survey of Netflix and Hulu users suggests most of us would rather be stationary:

Hulu Non-Plus users just hook their computers up to the TV; Netflix users are fine with Wii's low resolution.

Here, too, there aren’t many surprises, but there are a handful of items worth noting:

• Nielsen’s survey doesn’t distinguish between regular Hulu and Hulu Plus users. There are a lot more of the former than the latter, but you need a Hulu Plus subscription to watch content on non-PC machines. The easiest way to get around that restriction in the living room is to hook your computer up directly to your television, and 20% of Hulu users do just that. Attaching funky cables to your laptop may not be elegant, but it gets the job done.
• The Wii is generally seen as a sub-par media streaming device, particularly compared to Xbox or PS3. It outputs video at lower resolution, can’t play DVDs or Blu-Ray, and has far fewer streaming video options. But it’s inexpensive and popular, so 25% of Netflix users don’t seem to mind. (Disclosure: I watch Netflix on my Wii.)
• Also, 3% of Hulu users say they watch Hulu on Wii, which isn’t actually possible, at least without some hacking, even with a Hulu Plus subscription. What’s up with that?
• It’s a survey, not a census. A Nielsen spokesman told me they sent out e-mail invitations to a standing panel of 200,000 internet users, and 12,000 of them participated in the interview. I requested a demographic breakdown to look for patterns across age, gender, geography or household size, but Nielsen wasn’t able to pull that information out of the full report prior to this story. My suspicion is that people who are younger, ride trains and/or have small children might be more likely to watch video on smaller mobile screens.

It also wasn’t immediately clear to me how this data might be used, other than as material for a trend piece by tech writers like me. The business Nielsen is best known for is providing data to advertisers and television stations to help them price and target ad buys. How does that translate to post-PC/post-broadcast video?

“As advertisers, broadcasters and over-the-top providers like Netflix and Hulu are looking at how to address and price new platforms, it’s really important to understand how users use and think about each platform,” Nielsen VP Jon Gibs told Wired.com. “When consumers are looking at content on Netflix, they may be watching TV content, but they may not think about it as TV content. So how does that change the relationship consumers have to that content — as opposed to Hulu, which is much more clearly seen as a place to watch TV programming, often from just the night before?”

Part of the reason Nielsen is conducting these surveys, Gibs says, is that the entire television industry, including Nielsen, needs to philosophically decouple television programming from television hardware. This partly means thinking of TV content as something like a self-identical liquid that shapes itself differently inside different containers, but is always still television: “When we think about the way we produce currency in the market, the way we produce measurements, the way we look at impact, we have to look at all of those things as television content.”

It also means thinking in more complex ways about the containers themselves. “The big box on the wall isn’t just television content,” says Gibs. How do you evaluate something like ESPN’s new Xbox Live programming, which allows you to watch multiple games and follow data simultaneously? It can be done, Gibs thinks, but demands a different approach than ESPN through a cable box.

Instead of just thinking about a single screen, or even three or four, Gibs says (half-joking, half-seriously) that “we need to move to a mathematical model with variables like length of ad, resolution of screen and distance from face.”

Add in a fudge factor for bandwidth quality and an index for total number of hardware buttons required, and I think he just might be on to something.