The developers of the project are well aware that it is possible to perform all of the functions in 1.0 using existing stand-alone technologies such as Kerberos and NIS, but they are also aware of the significant challenges these ad-hoc solutions pose: particularly significant is the difficulty of migrating from these solutions, as well as the expertise required to set-up and administer the system. To combat this significant effort has been put into documentation on how to link with lots of different external systems, and there is also a strong focus on ensuring that it is not just the code that is free but the data that the system generates. To this end, FreeIPA aims to use standard formats such as XML/RPC as well as making the data easily accessible within the system. There are a number of existing commercial products such as Symark's Powerbroker, but these are often expensive and are built around around proprietary technologies, re-inventing the wheel when the building blocks of these systems already exist. By building on existing free software solutions and focusing on allowing the companies to own their data, FreeIPA hopes to provide a compelling reason for migration.
It seems there is a lot to look forward to in FreeIPA, even if the current implementation is somewhat limited. If you are interested in trying FreeIPA for yourself, the website has detailed information on how to install and configure it on Fedora 9, along with information on how to configure client systems. And if you are interested in working on the development of FreeIPA, and perhaps guiding its direction, you may be interested in discovering their community tools such as the mailing lists and IRC.