How does science work, really (2)?

I have seen two scientific problems up close, that suggest science doesn't work the way scientist or technologists think it does.

Let's begin with a major $5 million research project by Donald Glaser, a Nobel physicist who became a biologist.  Glaser found that bacteria changed their shape in reaction to their environment.  He used a mechanized system to grow and identify a million bacteria at a time.  He was able to correlate the bacterial shape with solutions of hundreds of fluids from diseased patients.

In one experiment he used urine drops from 30 members of a Stanford team (Glaser's lab was at U.C. Berkeley) and detected over 31 common diseases. 

 Glaser never got re-funded by the Federal agency that supported him.  Why, because he found a way to detect disease but no one knew the connection between the effect on the bacteria and the emergence of disease. None of the 30 men showed signs of disease. Also, the machine was too expensive to produce in abundance.

Next, I met an interesting scientist at a Tokyo resort who invited me to his lab at Hitachi.  When I went to the lab he showed me a newly operational x-ray scanning microscope operating at a wavelength previously never used.  The lab had many images of different molecule sizes from metallic elements. The problem was the images didn't connect with any known atomic surface patterns.  Basically the question was 'what are we seeing at the atomic level?'  Comparisons were made to other atomic level microscopes, but there was no connection. The answer was not found over the next few years so the research was terminated.

It looks to me like technology is regularly investigating the Real World and finding unexpected phenomenon for which the preceding (intermediate) steps of research don't exist.  So the research remains in limbo.  Much like the work of Mendel who applied math to pea reproduction with astounding results but wasn't understood for forty years.