The goals of this project were to create a Gallium3D PowerVR driver, port OpenGL 3.0 Mesa to this library, investigate possibly using LLVM to take advantage of multi-core SoCs such as the ARM Cortex A9, and to potentially use some sort of DirectX accelerator for use under Wine via Gallium3D (such as the Direct3D 10/11 state tracker).
The goals are quite ambitious, especially considering none of the core Mesa / DRI developers have been working on open-source mobile/embedded GPU support nor are the hardware vendors, but the Free Software Foundation found it worthy enough of classifying it as a high priority project. PowerVR graphics are found in many of the embedded ARM devices on the market currently.
What was achieved with this work at the time the FSF made the announcement was just reverse-engineering the USSE op-codes and the assembly language. But that was about it with no prototype driver or any planned road-map for making this open-source PowerVR driver by the community a reality.
Sadly, there's been zero progress on the matter. Today I asked Luke Kenneth Casson Leighton how the work was going. Luke is the developer that started this proposal. No one else has contacted him about contributing and his first priority is on receiving money before doing any further work.
The Free Software Foundation does have its "High-Priority Projects Fund", but that doesn't seem to be doing any good in this case nor have their tagging of it as a high priority project resulted in any development support from individuals or organizations. This though isn't a total surprise since back when the FSF declared open-source graphics drivers a high priority item, you didn't find significant contributions from them towards the Nouveau project or Radeon reverse-engineering, etc.
Meanwhile, one of the other community projects that's focused on creating a reverse-engineered driver for Samsung's S3C6410 embedded graphics processor, is moving along. The gles6410 driver is aimed at providing an open-source OpenGL ES driver for this Samsung SoC used on a few mobile-phones and tablets and it continues to receive new work as shown on its Google Code page.
Some of the other High Priority Free Software Projects are GNU PDF, Gnash, Coreboot, a free software replacement Skype, all free software video editors, a free Google Earth replacement, GNU Octave, a free OpenDWG library, reversible debugging support in GDB, free software drivers for network routers, a free software replacement to Oracle Forms, and free software automatic transcription software.
At least some of these priority projects are moving along like Gnash and Coreboot. Gnash though has been a high priority project by the FSF for years and it's slowly but surely moving along (though the Lightspark project as of late seems to be doing a better job at supporting the latest Flash/SWF features). Gnash still hasn't found its way into many Linux distributions as part of the default package installation for providing open-source Flash support. Coreboot is also still largely a hobbyist project at this point with no tier-one IHVs yet even shipping Coreboot-powered products.
Other projects like creating a free software replacement for Skype that's comparable in terms of features and support, creating an open-source replacement to Google Earth, and automatic transcription software like what Google has within YouTube, all seem to be pipe dreams and will likely be that way for the foreseeable future.
Meanwhile, some argue that the Linux Foundation is only good for issuing press releases, organizing events, running contests, doing interviews with the press, and other promotional/managerial roles, but not for doing things like delivering on their OpenGL patent investigation and other such technical roles. On the same note, the Free Software Foundation may continue to steer GCC and other projects and maintain other roles for advancing free software, but their high priority project campaign doesn't seem to have much of a measurable effect.