Nevertheless imho there are some things to be desired:

• The video clip should be self-contained, even if taken out of its framework (youtube). This means it should at least contain the basic data in the intro (event, date) and maybe the trailer (producer, license). The clip may be used in a different context (sent to someone on a CD, shown from harddisk, downloaded by some) and without an internal description in may become undecipherable. Who said this? When and where? In what context?
• When people are speaking for the first time in a clip, there should be a subtitle containing at least their name, maybe their function and origin (e. g. "Andrius Kulikauskas, participant from Lithuania") . Someone looking at a clip should be able to cite: at the GRUNDTVIG workshop at Vienna in January 2010, Andrius Kulikauskas suggested this typology ... for VB events. Videos are documents.
• Headers and trailers should be reduced not to marginalize the content of the clip. 3-10 seconds might be nice. Some trailing text can be implemented as a still frame, the video stops anyway. Time is too valuable to waste it. If you bore the audience, it will turn away.
• The clips should be content-oriented. Some are better in this regard, you can learn a bit from them (e. g. Andrius's typology). The clip of the 22nd is a bit weak in that regard. It's acceptable to capture the atmosphere, show people joking. But the three lecturers of the day all carrying the name "Thomas". Ha, ha !? We see one of them, but none of them speaking. So what is the message?

Be aware that your enthusiasm of having some video product created may not transfer into an enthusiasm of an audience that is used to Hollywood blockbusters and TV soaps. If we can't reach the technical perfection, we should at least aim to be authentical in our communication and valuable in our knowledge content.

Of course, everything could be underpinned with pattern theory. E. g. patterns are about transformations and suggest that you are aware about the context the initial and the final state, the pros and the cons. So you ask, why things are created the way the are and whether this fits the needs. E. g. you could look at the interface of the video as a BOUNDARY property increasing its contribution to living systems in the sense of usability (framework of youtube or DorfWiki, intro, ...). E. g. one could look at the relationsship of content and intro/trailer, and the relationsship of atmospheric and work information, under the property LEVEL OF SCALE (I prefer the term PROPORTION) and SYMMETRY. This would lead to a much deeper analysis and adaptation. But at that point of development this hardly makes sense.

FranzNahrada: Thank you. Good points! We will discuss them.