When we were evaluating the Lunchbox, I watched a couple of videos on my computer. I have the sound card patched into a "real" stereo rig in my office: the line-out is plugged into an input on a stereo receiver, which in turns feeds four two-way speakers and a sub-woofer. Even so, the sound was inferior: no doubt recorded via a camera's built-in mic, and ruthlessly compressed to get the audio and video to stream. I like to expand the video to fill the 19" screen on the computer, and, when I do so, the low quality is obvious through pixelation and jerky movement. The audio suffers the same problem as the video: what remains of the original sound after negotiating the limited frequency range of the recording mic, becomes squawky and coarse from the data compression algorithm. Speaking as a long-time recording musician, I have never been happy with the sound of mp3s at their best; as heard on computer-based sound clips, they may be at their worst, especially when they are sharing bandwidth with jumpy video.
In other words, the audio/video of the Lunchbox made me think it still had a lot to prove, for me to accept it as an amplifier I would use. And I believe that "sound clips" would suffer the same problem: without disrespecting the capabilities of my peers in this review, we would have at least four microphones choices, four signal chains, and four variations on the best way to EQ; all of these coming from four rooms with widely differing acoustic qualities.
The end result would be essentially useless, and my remark at the end of my evaluation section would still hold true: "This is an amplifier with a decided character, and you, it and your guitar may or may not find mutual happiness." Sound clips would not illuminate that statement any better.
"Digo: 'paciencia, y barajar.'" -- Don Quijote de la Mancha, Part II, Chapter 23