In a recent article in EETimes ('Pentagon urges 'relevant' R&D' , 23rd October 2006, pages 1 & 80), George Leopold reported on a marked shift in US federal R&D funding away from high-risk, high-reward defence research towards technologies needed to counter immediate threats.
There appears to have been some indications of this emerging trend in recent years, but this definitive position could have a significant impact on the approach taken in the development of many programmes. During the Cold War era, the nature of the threat and an enemy’s capability could be predicted, and the development of new defence technologies could be planned and executed over quite long development timescales, sometimes even spanning decades in the case of projects such as the RA-66 Comanche helicopter, F-22 Raptor, and the Eurofighter Typhoon aircraft (wikipedia). However, in today's uncertain world, new types of threat are emerging, and armed forces are now having to adapt rapidly to defend against them in-theatre, and it has been argued that some programmes have been cancelled because they cannot adapt to the emerging threat ('Army cancels Comanche helicopter' (CNN , 23rd February 2004).
I discussed the role of software in the changing response in a recent article "Achieving Air Superiority" (Defence Management Journal, Issue 33 June 2006, pp83-84), but there is of course the wider issue of the changing response at the platform level.
I've noticed that some programmes are already proactively using technology insertion / refresh cycles during long development phases to address technology obsolescence issues, and this approach could conceivably be extended to enable programmes to cope better with changing operational requirements. A simple example is unmanned air vehicles (UAV), which can inherently cater for different types of payload (e.g. SAR, Electro-Optic / Infra Red sensors, etc.).
However, many platforms are extremely complex, and the technology insertion / refresh cycles present their own challenges, and even when using Commercial-Off-The-Shelf (COTS) components, it's important to distinguish between COTS open architectures and COTS closed architectures. The former, uses a standards-based approach, using commercial standards (IETF, ARINC, POSIX - see Alex Wilson's recent blog POSIX profiles) and military standards which are widely accepted by NATO members (e.g. NATO STANAGs), and provides the basis for portability and interoperability. The latter, well...it doesn't use this approach and has the potential to frustrate future technology insertion / refresh cycles.