In terms of elegance, systems integration ranks very low, not much above the proverbial process for making sausage, and the result, at least for people who enjoy sausage, is not as tasty. The main issue is that systems integration represents an intersection between technology and business. The work is likely to involve multiple organizations whose goals may be in conflict. The outcome, if successful, is always a compromise. This outcome always destroys the beauty and simplicity of the constituent technical components, at least from the advocates for those constituent parts. There is a lot of drudgery involved: many meetings and repetitiousness due to the desire for predictable outcomes without hitting cost overruns. I'm sure qualifying someone as a "brilliant systems integrator" will only result in a healthy laugh. Somehow, this assessment is not in the same league as "brilliant scientist."
However, and the big however, is that systems integration constitutes an essential activity to any society where technology matters. It's the last step where a number of technologies come together and their value to society is realized. Systems integration is what placed man on the moon almost forty years ago, and our inability as a society to do execute on systems integration today is what's preventing us from doing what Neil Armstrong did on that fateful day of July 20, 1969, in spite of the far more advanced technologies and component subsystems available today. Systems integration is what allows the bits from this article to flow to your computer through the Internet. The Internet is "flawed", rife with viruses and malware, broken connections, and if you are using a PC, you probably don't want to look inside. Yet the whole system somehow manages to work.
Robert W. Lucky, a regular contributor to IEEE Spectrum reflects on this dynamic very succintly: http://www.spectrum.ieee.org/sep06/4387.
On the flip side, systems integration brings in a holistic view in addressing, the only hope for addressing some otherwise intractable problems. I have found the use of anti-spyware programs to get rid of malware quaint and peculiar. This approach is certainly beneficial to anti-spyware vendors. IT help-desk support may recommend it because that's what they can do within their mission. However, the big picture view of the problem in this context has notoriously absent in the larger industry debate. A new realization may be in the offing: Richard Stiennon, a columnist for NetworkWorld reports in http://www.networkworld.com/columnists/2006/112706stiennon.html?fsrc=rss-security about the economics of cybercrime. Spyware, and its close kin, spam and phishing programs exist because there is a strong economic incentive for malware writers to do their nefarious deeds. Hence the only hope to address this problem is to take on a systemic approach. A technical solution alone won't work. An effective strategy will be of necessity multi-disciplinary that includes understanding the economics behind this behaviors and will require the collaboration of corporate and governmental organizations, and legislative and law enforcement at the city, state, federal and international levels. There will be clashing interests, hidden agendas, nasty politics and what not. To be realistic and practical, these items must to be seen as part of the system integration process, not as barriers to the process. At the end, one hopes that in spite of this ugliness, the greater good will eventually prevail.