I offered both as solutions. Adjustment of a valve (and slight height change for really fine tuning) of the speed. Because adjustment of a valve can be only so accurate! I have found that an aquarium air valve is very good for just what you suggest. Mine has 2 controls for the valve. I get one just right and use the other to start it in the morning. The pipe gets airlocked at the end of the day when the bucket emptys. I get water flowing in the pipe and then close the first valve and know that the second one is set up correctly from the previous day. This waterclock system is the purist system. But you do have to reset it manually in the morning. This is difficult in cloudy days. A solution is a watch and hours scribed on one of the "gears". Then it would be possible to reset it even on cloudy days. I did not do this. In the other clock based mechanism, some of the water was not dripping (rather it slid down the pipe). The flow was not gravity dependent. It was totally dependent on the clock "wheel". I believe the clock based tracker is preferable because it is time based, repeatable, and to reset it in the evening, you just lap the string round the clock wheel twice and refill the top bucket. I find 2 buckets much better than one bucket. The ideal bucket for the project would be a section of a pipe (Same diameter top and bottom), but we use what we have! Unfortunately, I dropped the clock shortly after making the video and ruined the internals. Perhaps someone can come up with something more sturdy? Hopefully there are different makes with different strengths. Both are no more! I had to make space for my heliostat experiment. I will be using a dripper tracker (Probably clock based) as the timer when it is up and running, Brian.

There is one now on http://solarcooking.wikia.com/wiki/Tracking at the bottom of the page. It is by a guy called Michael Murphy who did a really nice clear job. (I am not capable of neat work)
Thank you Mike!

I will do a schematic in a couple of days. I put the video up (even though it is awful) just in case i dropped dead or got hit by a bus.
I have another tracker that is potentially much better on the solar cooking dot org wiki somewhere. It would be controlled by the hour hand of a clock and it is probably more reliable and more accurate than this one. (I cannot make one yet because i have to find the right clock). http://solarcooking.org/ has this video as recent news and another link is on the word tracking. (The clock idea is the second tracking device on that link). Having done the dripper experiment, I see that the clock would be far better attached to a tube OUTSIDE the bucket. The clock tracker for equatorial mount would be as good as the best of the pv trackers in my opinion.
The only special instrument you need is the clock. You could make the rest of the tracking equipment like in my video! (from old junk!)
Brian

Tracking will work for solar ovens too. (just mount it on a lazy susan). Please stay on topic. The topic is cheap easy tracking. Check out commercial tracking sites for how much tracking improves solar panel performance. If you can get this performance boost for less, then why not? Brian

It is good but no low tech trackers.