Blackberry was excellent in the corporate market for users, although the implementation of such was a bit of a hassle for the IT bods. My guess would be firstly adopt touchscreen technology and ditch the physical keyboard earlier, and secondly to adopt Android as the underlying operating system earlier for broad application compatibility, but for RIM to heavily customise and build on top of it to make it more suitable for the enterprise market that RIM used to make good money from, creating as "RIM-Android" flavour if you will as opposed to "Google-Android". Google is only now really making significant progress in enterprise functionality and addressing corporate concerns over the viability of open source software (whether you agree or not, enterprises and security people have generally been less then favourable to open source software); provisioning and management have been of varying levels for a while, and the OS/OEM split makes long term support more difficult.
A unified RIM-Android offering would allow long term support for devices (for patching and functionality), guaranteed fixing of security flaws, wider application compatibility with a "RIM Store" which developers can upload apps in to (and for which apps would broadly be cross-compatible with Google-Android, whilst also providing a better enterprise targeted version of an app store), and retain the business functionality RIM was well known for, all without having to maintain their own OS at heart.
Whilst RIM did this in later years, my view is that it was too late, and it was only a half-arsed implementation of Android and what it could be. Enterprises needed better functionality, management and security built in to the device, and will pay for it.
I also agree Blackberry tried to play with the big boys in the consumer marketplace and got burnt; they need to recognise what they are good at and build on that rather then the consumer space which is tiny profits on huge numbers of devices.