Subjective Conclusion
The foregoing analysis, though neither objective nor particularly rigorous, leads me to conclude that the variability in program outcomes is indeed related to the degree of project alignment with key leadership principles. In particular, I think that the two growing/sustained endeavors, namely NSFNET and Unidata, aligned more closely than the other two [7] with specific principles of leadership excellence put forth by Collins, Treacy and Wiersema. This suggests that NSF might wish to evaluate such alignment as a part of designing projects where collaboration is critical.
If NSF takes this step, I recommend gaining the assistance of experts like Collins, Treacy and Wiersema in developing an evaluation framework that is tighter than the one I have employed, eliminating redundancies and explaining the principles in terms specifically matched to proposal evaluation rather than the informal, reflective mode of my analysis.
[7] I recognize that comparing the outcomes of NSDL with those of the much older NSFNET and Unidata projects may be premature. Nonetheless, at comparable ages, I think NSFNET and Unidata exhibited greater evidence of agreement about the natures of the attendant collaborations, especially in relation to the central facilities. To the extent that my thesis is correct—i.e., that this lack of agreement is related to leadership shortfalls—then I certainly share culpability.