There are a couple of things that have been on my mind for over a year now. If I were working in an educational institution I think I'd pick one of them for a thesis.
The first item is flow equations as it relates to data. Think about it, electrical, gas, and fluid, for that matter probably traffic flow, all have equations and models for calculating flow. Take a look at what is in Wikipedia as an example: http://en.wikipedia.org/wiki/Groundwater_flow_equation. The need and even mechanism for calculating the flow is common. As I think about this in respect to data, it also has similar characteristics. It has to flow through different size channels and mechanisms that act as different size conduits and even different type mediums, it gets consolitdated, it gets dispersed, and even the data itself could be attributed as being of different viscosity. So, my thought is that there are ways to create models with which we could represent data flow much like fluid. I think this has similar, but somewhat different models microscopically (within a single application) or macroscopically (across systems, organizations, etc.).
My second thing, which is probably harder to nail down with facts, figures, and models is the concept of software genetics and evolution. Specifically, I think that software and systems have a genotype and a phenotype. Let's face it who imagines their solution imperfectly. When one starts with an idea, with a design, it is envisioned perfectly. At the point after it is imagined and starts to factor in reality is when the maturity and experience of the architect/developer comes into play. The alterations made to a design or implementation based on fitting it into its target environment represent its phenotype. In fact, the changes made under maintenance is its evolution which inevitably will get reflected back into new designs thus changing their initial genotype and then the cycle of life and evolution continues.
If anyone reads my blog and knows someone doing research in this area, post up, because I'd like to get involved.