I have a TG-6, and after much trial and error, I’ve now given up on using the GPS of this camera (or more properly, waiting for a GPS location to be derived, since I still keep it active just in case it actually gets a fix before I move on). It takes anywhere from ten seconds to over a minute for the camera to stop blinking the GPS symbol on the video screen, indicating it has a fix, even when the GPS “log” is turned on (switch on top of the camera). It takes even longer if you don’t have the “log” turned on.
It is much faster, and more accurate, for me to take a GPS point with my Garmin Etrex, and then manually enter it into the iNat location (there are programs to automatically associate the two, but I haven’t tried them yet).
But if you don’t have a stand-alone GPS, or don’t want to enter it manually, then you must wait until the GPS icon stops blinking before taking your pix. Only then will the pix have a GPS location in it.
Here’s an analysis I’ve done of the accuracy of iNat locations:
http://tchester.org/plants/analysis/iNat/position_accuracy.html
The main problem with camera and smartphone positions is that they don’t actually keep a running GPS track sampled every second. They take samples much more infrequently, so that there is always the possibility of an “initialization” error for each fix.