obobskivich
Headphoneus Supremus
No need to repeat louder what you already said clearly, I'm not deaf.
Weird, the quote won't grab the rest of your post...
Anyways, no it should not go to 32 mega-ohm.
http://www.youtube.com/watch?v=4HvHMIMYbBk (roughly what you're talking about?). Water doesn't out and out resist motion, it just doesn't compress. You can drop a driver in water and it will flap around until it shorts (and the short will be a crossbar and take the ampilfier with it). Adding mass to the cone will modify (stabilize) impedance - it's a form of damping (and it's not uncommon with very large subwoofer designs, and fairly common with passive radiators). Here's a really basic (and exaggerated) simulation for you:
As you can see, adding 1.5kg to the driver helps to knock down that impedance peak, and it influences driver excursion as well. It will change the FR too:
And it makes a lot of other changes to the overall sound of the system (it would probably sound like hell, but there are drivers or PRs that are robust enough to deal with 1.5kg); basically it would "flatten" the response by dealing with the impedance spike and making the driver's response more even, but it introduces issues with group delay (which goes off the plot and is probably what would contribute to it sounding like hell) and maximum output (it can't move as much air due to being weighed down). You can tame the impedance spikes with a Zobel network without limiting your peak output (as long as your component values are high enough), which makes the system more universally similar across sources with different Zsource values (TMK there are only two headphones in the world that do this: the Sony MDR-F1 and MDR-MA900).
More technical and on-topic:
http://www.sound.westhost.com/impedanc.htm
With headphones, this becomes something less of a problem, because good damping isn't impossible to accomplish, and very small drivers with very large motor structures are generally fairly well controlled to begin with (and that's most high quality headphones). TMK the maximum peak power ratings are always thermal limits, not Xmax limits, given that plenty of anecdotal evidence exists for ABSURD amounts of power being dumped into headphones short-term with no apparent damage (hundreds if not thousands of times their nominal operating range - conventional woofers will not survive that kind of abuse).
:EDIT
First reply lacked sim example.