On a large scale, solar tracking only makes sense if you are reducing the density of panel area to land area. When the sun angle is lower, the solar power incident on the land area is reduced. Trackers don't improve on that unless the panel area is nominally less than the land area. If tilted panels don't block each other somewhat, then at high noon you have gaps between the panels and no noon peak of solar power. The arrangement where tilting does have a benefit is an arrangement with less panel area and less power at noon than if the panels were at full land density.
If you have one big panel, tracking can make sense. But it is true this involves moving parts, a control mechanism (based on timing or sensing), less reliability, and greater maintenance.
The benefit of aiming the panel is lost on cloudy days. Then light comes in at all angles. This needs to be taken into consideration if you want solar power during these times. Certain multi-angle solar concentrators, such as the Winston Collector, take light from multiple angles and concentrate it. Photovoltaic cells also operator more efficiently at greater light+IR concentrations. And the multi-angle concentrators take in more light on cloudy days.
http://www.paws.dircon.co.uk/solar.htmhttp://www.appropedia.org/Understanding_Solar_Concentrators
