We use the JSON output writer in Solr because it was considerably easier and faster to parse on the PHP side - pretty much a direct pass to the json_decode function. The response was usually smaller in size bytes wise. It's also one of the most stable and used other than the XML output writer (for example there are PHP and serialized PHP output writers, but there were / are some issues with using them).
There can be downsides to using a non-XML output writer. Some request handlers / plugins may not generate data structures that can be always be represented nicely in a hashmap / associate array manner (e.g if they repeat keys in a key / value list). This is why there's a list treatment parameter in the client that tells solr how to map key / value lists into JSON objects or arrays, but its not a perfect solution. All of the "standard" plugins for the most part generate a JSON object tree that makes sense to traverse.
I personally think our API is easy to use, but i'm a bit biased ;) . I have not yet used the solr module so I can't really speak to it. Both clients have generated documentation (mine is included in the download). I also have the wiki with some example code and this google discussion group - Israel Epko (the maintainer of the solr module) probably has something similar, I happen to know he's active on the solr user and developer mailing lists.
I'd be interested in what decision you finally come to, especially if there was a problem in our client that made you turn away. Good luck with your project and I hope I've helped.
- Donovan