I have some standard GP-40's that do that also....as near as I can see, it's due to the power pick-ups on the frame, they act like springs and allow some movement of the body.....none of mine ,however, ''jiggle'' excessively and that might be a problem with your specific locomotive. I'm sure others will have a more definitive answer for you.

The issue of nails through foam  is why I suggested long nails of the brad or finishing variety.....I have several boxes of 1'',2'' and 3'' nails I use for nailing through foam,  these are not the ''track nail'' variety sold at hobby shops , they are nails for moulding or paneling, they are larger than the ''track nails'' so some pre-drilling of holes is needed, I like to nail my foam down rather than glue it [except in rare occasions where I use acrylic caulk to hold the foam to itself ....in a terrain building layer, as an example], so in the future I can remove any areas I might want to re-do  and I always nail my track, the issue of the slipping radius is most likely directly caused by a failed bond of foam to track because the foam doesn't  lend itself to gluing....reason....most foam [that I've used anyway], has a clear moisture barrier sheet over the actual foam and this doesn't allow the glue to stick very well.
   Glue with some nails to help hold the shape of the curve and weights placed on it till it dries thoroughly ...as craig suggested...should also work well, if you don't want to try to find long nails.

Several times I've seen folks say that brake fluid works well to strip paint from rolling stock....I've never tried it so I can't say yes or no that it works.
   I've also been told that Castrol Super Clean works also, it's availible in the automotive section of WalMart or a auto parts store...again, I've never tried it so I can't say if it does, in fact, work.
  But, I do know this, NEVER use paint thinner\laquer thinner or any petroleum based products on plastic.

Nails do work on foam...if you've put the foam on wood....which you should do anyway....you just have to use longer nails....and don't drive them down so far they dimple the track, only pound them in till they are SNUG....not all the way down to China....the forces exerted on a piece of flex track when you bend it into a curve will cause it to want to straighten itself back out [the way the ties are made is the reason]....so unless you want to use some heavy duty permanent type glue....nails are the best bet.  Small brad type nails work very well and you'll find them in several sizes are your local home improvement warehouse.

Started in 1978.....on and off in the hobby till 2001,  then I got back into it in earnest.

My suggestion.....don't use glue on flex track curves...use nails.....when you bend flex track you put pressure on it and if  if the glue isn't very strong or not completely cured...it will let the track move.  I always use nails to secure my flex track.

Yes....it makes a pretty good feeder wire,  I personally wouldn't run it more than 12 ft from a power pack or larger gauge bus wire [a bus wire is the larger gauge  wire some use to connect feeders on to for constant voltage all around their layouts...the bus wire is connected to the DC terminals on the power pack, a feeder wire is connected to the bus wire then  to the tracks, usually by soldering].
I use doorbell wire because it's easy to solder and relatively cheap, I've never had any trouble with it and I don't get voltage drops anywhere on my layout.

The reason those MP sharknoses are so plentiful, MP sold their old stock off at bargain basement prices....$2 each for the a and b units [on their website]. I got a few to use as yard dummies, I've been seeing them coming up for sale on EBAY for as much as $20 each.....AND SELLING!!!!!  good deal for some people I guess.
     I make it a point to check MP and other train manufacturers websites for just such deals, right now they're {MP} selling BUDD coaches for $3 each and a set of old time coaches for $9....those too are on EBAY right now, go check them out you might find a deal on something.

                    JM