Using the gadget for screening makes a certain amount of sense, but why remove the chain to confirm with a ruler. It's easy enough to pull the RD back to get a taut section over a foot long in the lower loop which you can then measure on the bike.

Remember that measuring stretch isn't a precise measurement of the chain's condition, it's only an indirect indicator, so serious precision is meaningless.

The best explanation of chain wear measuring tools I know of is this:
http://pardo.net/bike/pic/fail-004/0...easuring-tools

The bottom line is that he finds commercial chain wear measuring tools are conservative in their wear estimates. But also that almost all of them includes roller wear in their measurements instead of just measuring pin wear.
A tool series that only measures pin wear is Shimano's TL-CN40/TL-CN41 and the new TL-CN42.

A number of years ago at one of the industry shows (East Coast, Philly I think) I asked the SRAM (actually IIRC they were just Sachs then) guys much of this same question. Their answer was that the manufacturing tolerances were loose enough (with all brands they tested, not just SRAM) so that you'd get a range of "measurements" on even new chains along the chain's length. I know I've gauged (and this is what most wear indicators do) new chains and gotten readouts suggesting much wear.

Still we use chain wear gauges at work daily. To some degree I feel there is an aspect of judgment at play. This is why i prefer to use the same gauge each time I check a chain. Andy.