It is really good to see more and more companies offering services that can be embedded in one's own webpage. YouTube is probably the most known example, but the concept is applicable to many other types of services. WordChamp's new Reader API is a great example of this. Between their Web reader, their Firefox plugin and now the Reader API, translation is easy to get for learners, web surfers or content provider.
I believe the future of language learning methods is with the web. The standalone language packages (like Rosetta Stone) take too long to develop, are limited in the amount of material they can package and do not leverage community effort (which WordChamp does with user translations).
The biggest challenge with the web based services is to convince people to pay for them. As with many web 2.0 companies (e.g., MySpace, YouTube), WordChamp is currently free, but obviously this cannot continue forever. I hope they figure out the profitable and sustainable model soon, as I would hate for their service to disappear.
A problem specific to WordChamp is that they are trying to cover most of the languages of the world. However they only have language-specific grammar rules (e.g. conjugation tables) for some languages. This means that grammar specifics of other languages are not picked up and some words do not show the translation even though one exists for the same word in a different tense, conjugation or declension.