I think that an important point with regard to this matter is that its significance depends entirely on the service in question. Services that only require a one-way transaction will not have this as even a potential problem.
Equally important is the fact that the issue is entirely addressable by the design of any given service (i.e. by a coin, not by the Blocknet), and these services have not been designed/optimised for the Blocknet.
For example, if a service requires that coins be returned to the client, then they don't actually have to be exchanged, and so an escrow service might work better.
This would eliminate slippage and thereby present a competitive advantage to any service that offers it.
The point is that you're worried about something that is (a) readily surmountable and (b) hasn't arisen yet. You're pre-empting an issue and then worrying about it as if it's a real problem.
It isn't. In a free market of coin-based services, there'll be competition on a new level. Coins will have no problem inventing clever ways of rendering quality services and rising to the top of the pile.
i am just looking at the example you previously provided...which involved mutiple exchanges
ok we'll see...
edit: is there some kind of automatic (and possible trustfree) escrow possible with your design?
There are a couple of trustless escrow designs out there. I'm no expert on this, but one of the more interesting ones is Truthcoin. It's one of the most visionary ideas I've read in quite a while.
More pertinent to a simple service provider might be Blackhalo and
this concept. No doubt there are a lot of other ideas out there too.