Actually, trackers are pretty expensive in $/W, and this is even after you take advantage of the increased yield (you're paying money to avoid the cosine effect so you better generate more power than the cost of your tracker). If you're going to do this you might as well use higher yield panels, which again increases capital cost, thus...
The economics of PV solar went this way:
- First the panels were expensive so the cost of installation was not a big deal. Thus plenty of 2 axis (typically azimuth/elivation) tracker companies sprung up to optimize the produced electrons/m2.
- Then the $/W fell below a dollar (panels were so lucrative a huge amount of factory capacity came on line in China and drove the cost down, just like the DRAM business). Now the cost of installation (still a couple of bucks back in 2012 IIRC) was the dominant cost.
- At this pont the panels are so cheap that cutting installation by 2/3 and just putting in more panels was cheaper than a tracker.
- Plus trackers had op ex (maintenance) much more than a fixed installation
For a while single axis tracking was worth it, but the price of PV has come down so far it no longer matters.
There are specialized applications (mainly where space is required, or concentration can benefit in other ways) where tracking is worth it and smoe people are still at it. Since the tracking motor itself is expensive, one strategy was to make a robot that went along moving each panel one by one. (QBotix). I don't know how well that has worked out.