There are some interesting points made in this thread. A quick check of the rules of harness racing will reveal that there are two machines approved for the testing of TCO2. The Beckman Elise and Beckman DXC 600 (I think it is). For testing to be valid the samples need to be tested on one of these machines. Obviously it is not so difficult to have one of these machines approved for use (usually a few months for the rule change to be approved). These machines also mut be calibrated correctly for use. Outside of a certain margin the results have more margin for error.
Secondly testing on the night/day has been conducted in the past. This happened in WA a while ago. It caused a lot of problems during an interdominion and was scrapped after that. Multiple tests were run on this horse on the night as the first test was over the limit, which was 35mmol then. This horse was favourite for the series. It would not have been good had it been not allowed to start. But then how could it be allowed to start when it was over the limit?? A tricky situation to say the least.