Gegenzeit wrote:
That way white settles one group and builds a wall with the other. In the first diagram one stone remains completly undeveloped while black has sente an could pincer and push white against his wall.
In the first diagram, black's group has no eyes, and no territory. Whilst it isn't immediately very vulnerable, a few more white stones could significantly shift the position. On the other hand, even if the black group is already strong, white has territory but black only has potential. White even has the cutting point which is worth a few more points. Overall, white has chosen to take territory in sort-of-sente (the position isn't really finished, but now there are so many context-dependent choices that there is no clear next move).
In the second diagram, white settles one group for not many points (but it is very strong), and his other group is quite vulnerable. Black can immediately start pincering this group and making its life difficult. Its position is not bad, but it is in no way thickness, and white certainly didn't get a free lunch on both sides. I would say that white has chosen to really cramp black's shape, but in doing so has committed to supporting a not-completely-strong group. That can be good, especially if that group already has backup, but it isn't so good on its own as it might look.
Obviously neither of these things is 100% the case in every conceivable board position with these shapes on, but both general ideas turn out to be the case, and conflict with your own assessment.
Quote:
My guess would be that white would only play like in diagram 1, if c4 already has backup nearby. But, if the position depends on a stone that is not on the board, can it be considered joseki?
C4 does not need to already have backup, though it may be useful. Because black's stones are eyeless, they aren't quite so carefree as they look, and it is difficult for black to mount a really strong attack on white's lone stone. This doesn't mean that white should necessarily immediately be trying to pull it out and save it, but it does mean that black can't try to enclose it on too large a scale (or it will just live too easily locally), and black must be wary throughout the game of the possibility of it escaping in an annoying way.
A lot of this stuff is not obvious because it's really difficult to imagine how the game might be progressing in even just 10 or 20 moves. I encourage you to try these shapes for yourself, and guarantee that your opponents will frequently be able to make your positions look much worse than you thought they were.