The problems are economical and priorities.
The cost per pound to put satellites into orbit is large. There are lots of arguments on how large, but there is no question that it is large.[1]
The priority problem is that unlimited comms and convenient navigation are only useful after a lot of other things become a reality.
Unlimited range comms will only be useful when moving about the surface is (relatively) fast and (relatively) unlimited. Navigation is the same - it is only necessary when you get beyond visual range.
Navigation can be done using (a) eyeballs, (b) inertial navigation and (c) the stars. Eyeballs are good for short distances. Inertial navigation is good for moderate distances. Stars are good for very long distances. Navigation is best done in layers: inertial (and/or stars) to get you close enough for eyeballs to work.
GPS on the earth is (historically) a very new way of navigating. It is very convenient, very accurate, and can be used with almost zero knowledge and skill, but it is not necessary. Navigating using the stars has been used for thousands of years. Over time, improvements to time measurement, mathematics, angle measurement (sextant), and star mapping made the navigation more accurate.
You only need inertial / star accuracy to a few miles - visual range - in order to successfully get to a destination. After all, Mark Watney was able to navigate successfully to the MAV using just Phobos and his eyeballs. ;-)
[1] Ref: https://www.reddit.com/r/Mars/comments/42rs7k/cost_per_pound...