There is of course, the issue that RIM already has it's own software platform, the BlackBerry OS, which has been under development for about 8 years.
While Android is Linux based, and Blackberry is completely proprietary, both share quite a bit in common from the developer perspective, in that the applications are written in Java. There are of course some religious differences as to how Java is implemented on both systems -- BlackBerry uses a licensed derivative of the Sun J2ME JVM and version 4.x implements a subset of MIDP 2.0, whereas Android's Dalvik is an Open Source re-implementation of Java that uses a unique Google-developed bytecode that is incompatible with that of J2ME, so it cannot be certified as "true" Java. Nevertheless, from a developer perspective, the two systems are very similar in terms of skill sets that are needed to create software that would run on either device.
Still, in order to develop Android and Blackberry apps today, developers need to maintain separate code bases and separate developer platforms. However, if BlackBerry and Android ran on the same JVM, they could in fact share the same developer environment. Developers would not need to prioritize which platform to develop for -- their application development target would in fact be the same.
How could this be done? Either by porting Dalvik natively to the BlackBerry OS, so that Android apps could run side by side with conventional BlackBerry apps and eventually phasing out the licensed Sun J2ME JVM over time, or by having RIM move to Android and port all of their enterprise messaging/calendaring integration services for BES (BlackBerry Enterprise Server) to that platform, adding any branding and UI customizations required in order to make it more "BlackBerry-like" and including the legacy J2ME JVM on the device to provide legacy app support during that transition period. In terms of level of effort, the second option would probably make a lot more sense.
There are a number of reasons why RIM might want to consider abandoning their own in-house OS for Android. For starters, both RIM and Google are facing three competitors that are encroaching on their space rather quickly -- iPhone, Windows Mobile and Palm webOS, and to a certain extent in the European and Asian markets, Symbian. A strategic alliance between the two where BlackBerry becomes the premier mobile Android platform for enterprise and heavy messaging users and would include tight integration with Google's web services would send shockwaves down Infinite Loop and Microsoft Way.
A RIM/Google alliance would not preclude the existence of other Android devices on the market, such as the T-Mobile G1, but presumably only RIM Android devices would be BES capable. RIM could also run their own Android store with apps that take specific advantage of RIM Android devices, maintaining the unique value of what makes a BlackBerry a BlackBerry.
RIM could also take advantage of Google's massive datacenter infrastructure, and enlisting Google's help in order to provide redundant NOCs, as opposed to the single NOC in Ottawa which they maintain now. This would render BlackBerry system outages a thing of the past, or at least much more infrequent than they occur now.
Would a RIM and Google alliance make sense and result in a "Perfect Storm" for their competitors, or are the two companies incompatible? Talk Back and Let Me Know.