I don’t think it’s possible to measure the exact delay, but it was designed the way to make it no more than several nanoseconds. But there are other factors depending on hardware setup, which are out of our control. For example, cables and their condition.
Reach RS2/M2 and Reach RS+/M+ use different chips because the first ones are multi-band and the last ones are single-band.
I saw your other thread. Tell us more about your project! Why is the delay such crucial?
NMEA was not built for speed, it was designed originally for big boats. Its lightspeed quick compared to 12knots.
Short cables, and high baud rates are best for delay reduction. The faster the data is received the faster you can use it.
The pps will only line up with the x.000s messages. Any measured difference will be processing delay at both ends and exceedingly small. Your pps detection voltage threshold of the rising edge on the receiving end will also matter.
There are very few new applications for pps, it was used to sync slow hardware clocks originally and keep them from drifting.
If speed is really an issue a binary like UBX will process much faster, because its much smaller.
But with either NMEA or UBX most of the communication time is idle anyway.