170 peak and 140 average means it jumps up to 170 and drops close to 110 during the first 30 seconds. How low it actualy drops depends on how soon it reaches the peak of course, the above is just the simplest guess.
The average power is still good, but there are lots of high peaking and low dropping CNI lasers lately. It is actually better to have a laser that peaks a bit lower, but has an average not much far away from the peak.
I have an "X-"150 (directly from CNI), that peaks between 176-179, and averages at 168-172 by Nova standards
(one 30 second measurement), but has a sudden drop to 135mW at 35-40 seconds, so the one minute average is much lower.
What the hell is CNI doing lately? I want to buy a 400mW laser with TEC from them, but they can't guarantee it will stay above 400mW for more than 3 minutes at a time (unless mounted to an external heatsink, so the TEC has somewhere to pump the heat)...
These high efficiency crystals can be very unstable with changing temperature and even have strange ideas sometimes (like getting tired). If the manufacturer would throw every imperfect set of crystals away, they would make less money or have higher prices, but didn't they used to be better?
Or could it be, that they started cutting corners, to make more money?
This laser was probably supposed to be a 150mW, but wasn't stable enough to classify as that by Nova standards. Could have classified by CNI standards, as they only promise 17 seconds of above rated power. In some cases these lasers can actually work better with a slightly lower current. The peak is lower, but the average becomes higher, and the sudden power drop can even dissapear completely. Too bad they are near impossible to take appart without damage.
The drivers in these are really amazing tho. I have an X-75 in pieces, and just tested the driver, and it only needs 0.39V on top of what the load needs, to keep a perfectly constant current. Almost as low dropout voltage as the AMC7135, but adjustable with a pot!
Have fun with your new burning machine! 
