Actually, it DOES work quite well (which is a huge surprise to me).
I'd trust it as much as a $300 Snap On unit. Not as much as an $8000 Alber Cellcorder we use at work, but it really is a pretty useful tool.
It's interface *****, but you can't beat the price, and you can get usable data from it.
Whoa nice! I need to get one!
I got mine about 6 months ago, and tested about a dozen batteries of different sizes and conditions with it, comparing the results against the Alber Cellcorder's results in a few of those cases.
If you test a battery 10 times, you're likely to get 10 different numbers, but the relative distribution of the numbers was reasonably tight (I'd say around 5% variation). Of course, the Cellcorder's variation was much less, but the variation in the HF unit was small enough that you could still get good information from it).
The actual internal resistance numbers reported by my HF unit were around 20% less than the numbers reported by the Cellcorder. This could be because it uses a different method of measurement (although it really appears to be a 4-wire resistance meter), or perhaps a calibration error.
From other testing, I've found that the difference in readings between two Cellcorders is within the normal variance between two readings (that's pretty amazing actually). I have not compared a second HF battery tester to see this difference, but I suspect that this would be it's biggest failing.
The way to get the most out of a meter like this, is to keep a record of a battery's history (and in the case of this HF meter, I'd trust it most if the same meter were used for every reading). You would want to charge the battery fully, and then take a reading when it is installed, and record the numbers. Now you have a baseline you can look at, to see how the battery is aging. An accelerating decline in CCA (or a CCA that's well under the nameplate rating), is a sign of imminent failure. This is harder to predict if you're just handed an old and untested battery, and have to interpret a single reading.
So, where does it lose in quality:
Well, for one thing, it has no battery of its own. I actually kind of like this, because there is no battery to replace, however, it has no memory, and every time you attach it to a battery, you need to wait for it to "boot up", and then you need to go through the annoying startup again.
For another, the clamps are really cheap plastic, with almost foil thin metal alligator contacts. You really need to adjust them a bit to get a good connection, and it is critically important that BOTH sides of the clamp have a good connection to the terminal (this is because of how a 4-wire reading works). It is supposed to warn you if the connection is iffy, but that warning is itself somewhat iffy.
The screen itself is very easy to read, and the back-light is a nice touch.
It is best to enter the CCA number when it starts up. It isn't absolutely critical, but large differences do appear to change the outcome a little bit (possibly because it would make different inferences about charge state, and that would change the final CCA calculations).
Entering the CCA number is the most annoying part. They really could have done this better.
So, final vote: PASS
For taking a reading on a battery, it is very good.
For annoying the user, yeah, it is good at that too.
Will the Snap On tell me more about the state of a battery. Probably not (although some Snap On meters will also test the charging system, which this will not). Would the Snap On be easier to use. For sure.