Hi, sure any help is always welcome.
before you start coding, I was asking to why? as applying a filter is usually as good, it will result in the same functionality.
The reason is just for performance reasons, as if you don't need to search in multiple shards the query can open a smaller index than the big one, still this performance improvement is noticeable only with extremely large databases, as indexreader's are pooled and optimized anyway;
another reason might be legal requirements, but are you needing that - it's for those cases people is enforced to use a database on a separate machine for each customer - not a common use case.
Anyway, you're welcome to help out implementing it as I'd love to have the feature.
Start with this reading: http://community.jboss.org/wiki/ContributingtoHibernateSearch
Then subscribe to developer's mailing list: http://hibernate.org/community/mailinglists hibernate-dev
And send an email there, just to say hi and that you're interested in contributing it - I and others will send suggestions and directions from there, as these design decisions are better discussed in team and other developers are unlikely to see this thread. You can also discuss it on the JIRA itself, follow the link I sent in previous post.