Hurricane intensity change is composed of external factors, many of which we have a handle upon, and internal factors, many of which we do not have a handle upon.

Start with the internal factors: eyewall replacement cycles (which we can see happening, but cannot predict; generally, as an eye tightens, the storm strengthens, only to weaken as it begins to replace itself), inner-core dynamics (such as vortex Rossby waves, the breakdown of the eye into the mesovortices seen with Isabel and other storms, and so on), convective hot towers and their distributions (can enhance a storm's winds & impacts locally for short periods of time, but communicate their energy on longer time scales towards the storm, both negatively and positively), the efficiency of the storm, and other such factors. We can analyze them after the fact, but don't know enough about them and their impact upon intensity change to accurately quantify them. We do know, however, that they likely play a substantial role in major hurricanes (in particular).

Then you have external factors: upper-level conditions (whether shear, temperatures aloft -- goes into maximum potential intensity, or the position of upper lows to enhance outflow/provide a means of enhancing a storm's secondary circulation...a fancy way of saying enhance the inflow and outflow and thus enhance the storm), sea surface & near-surface temperatures, interaction with land masses, dry air in the atmosphere becoming entrained into the system, gravity wave propogation, any sort of longer-term climatological factors that may influence the storm, and so on.

In general, warm ocean temperatures, little shear but a means of ventilating the storm system, little interaction with land, and no dry air (dry air is bad -- helps to increase the potential for downdrafts in the mid-levels of the atmosphere, leading to convection dying out and a cut off of the means of energizing the storm) are all favorable, with the opposite being unfavorable. Amongst the inner-core factors, a contracting eyewall is one of the better-known signs of an intensifying storm, whereas the completion of such a cycle is a sign of (usually) temporary weakening.

I think that about covers them all. It'll be interesting to see where Dennis goes from here, both track- and intensity-wise, particularly with many of the parameters for sustained intensification present within its environment. I don't have the list of the 7 rapid intensification parameters off-hand, but most of them do deal with many of the factors mentioned above, plus what the storm is already doing.