3-4 invasion: bad result  

I think the first diagram is pretty misleading, assuming that a move at is any good. The reason for which a move at is bad for White is the same reason is just as bad. If white plays creating a wider side Black punishes it by playing at (in the diagram) increasing his influence over the wider side and getting leverage on a part of the board that was made more valuable by White's mistake. However, the problem with the move order change in the actual diagram is the opposite, Black creates a shorter side with , immediately making both and inefficient. The correct move by White should be one that ignores the short side, like the 4-4 point in the top left. This is proven true (according to ps.waltheri.net's database) in that only 30% of pros playing Black in the resulting diagram win.

Alternative  

The result is that the left side and the top side are equally valuable, making them miai. But black has created a shorter side, so after White gets either of the more valuable sides he can approach the lower right in a way tht will leave Black and White with equally valuable top and left, but Black's right side will be less valuable than White's lower side.

Potential follow-up  

It is very important to choose the proper joseki however, emphasizing the difference in the side's value and forcing Black too keep the shorter side, which Black of course should not allow. However, this shows the problem with in the first place, it serves no purpose to play prioritizing influence over a side you do not intend to keep, for which I consider that such a move is a trick play with little strategic depth other than trying to induce a poor position through a timing mistake from White.

is fundamentally wrong  

In this variation White fails to capitalize on Black's mistake of , by allowing him to make an enclosure, more so, an enclosure that devaluates white's lower side with it's influence.

Consider  

Playing like this, White actually tries to make profit from the mistake of , instead of passively letting Black fix it. It is clear that white is left with the (however slightly) wider side, hence, Black's stones are not working as efficiently as White's, hence, they are not efficient enough.

Pro play  

Pros will not play at a immediately, but will instead further capitalize on the mistake by playing around b, ideally ending up in sente to play a as soon as the exchange ends. this is to reduce the influence of the marked black stone, making it even less efficient.

Move order.  

As for the order of the moves, it is actually important. Kajiwara gives the example in The Direction of Play as a method to punish , by playing . Where White creates a wider side, and Black invests influence over it immediately to take advantage of its increased value. But playing at b after at a makes wider sides of the top left and bottom. With an odd number of wider sides and White in sente he should strive to obtain 2 of them and make Black end up with one wide and one short side. This is very different from The Direction of Play, in that the order played in the book is a legitimate way to punish a mistake, while the other is a trick play trying to induce a mistake from the opponent's fear of 4-5 stones.