NavList:

It’s funny, when I first started reading about GNSS, the word was that the signal was very difficult to jam, because it was ‘spread spectrum’ ‘Hedy Lamarr’ and all that.  Then someone made a jammer and found it worked, so poor old Hedy slipped out of the headlines again.  It’s true, there have been incidents, intentional and non-intentional, and in the UK it’s been thought important enough to issue ‘Notices to Airmen’ when jammer trials have been taking place.  However, it still seems to me that any jammer would have to be very highly targeted against a particular area or victim.  I suppose jammers could be used as a defence against GNSS guided weapons, but it’s important to remember that any transmitter can itself become a target, e.g Flensburg v Monica, Naxos v H2S.
I’m not sure about the military receiver argument.  The magic inside a military receiver just means it can decode military signals in real time when the civil signal is either lost or is deliberately slipping in a few porkies.  If an aggressor can overwhelm the genuine received signals, it doesn’t matter that you had the innards to decode them.

As I understand it, the eLoran argument isn’t just that it gives a potential aggressor two aids to extinguish instead of one, it falls more into tradition transmitter v jammer mathematics, so an eLoran jammer couldn’t be the sort of thing you could carry around in your pocket unless you could get very close to the target.  You would need quite a large jammer, and of course, big means time to build, build means observe, observe means counter.
Just to slip another fly into the ointment, what about a problem with the Air & Ground segments themselves?  See http://www.rin.org.uk/newsitem/4415/GPS-timing-problem. 
As ever, I think the eventual solution to many of these problems, PN if not T will be integrated systems e.g. combining the short term accuracy of inertial with the long term accuracy of GNSS or an astro tracker rather than using two different external radio systems.  DaveP