That's definitely one of the sections I want to make more intuitive in the next release of the wallet. Right now, you have to do it the same way as with Bitcoin-QT/etc - via the RPC interface. If we could have a 1-click (or in reality probably 3-click... but 1-click seems to be the buzz phrase regardless of how many clicks) way to import them, it would certainly make our physical distribution strategy more intuitive.
Here is how you do it now:
Help -> Debug window -> Console
If you have a password on your wallet (which I'm sure you do) you will need to temporarily unlock it first by typing:
walletpassphrase "yourpassword" 180
With the "180" simply being how many seconds you want it to remain unlocked for (just needs to be long enough to run the next command... I always default to 180 but it doesn't really matter).
Then you type:
importprivkey "yoursuperlongprivatekeystring" "label-here"
The label is optional, of course. If you are on a slow computer, you can also add rescan=false at the end of the previous command and simply let the blockchain do the work for you. It allows for a smoother import and I will run some testing on it but I may make it the default on the new wallet import feature with the option to check the "rescan" box instead. Just to clarify, if you don't want to do the rescan the command would look like this:
importprivkey "yoursuperlongprivatekeystring" "label-here" rescan=false
That's all there is to it... but obviously not very intuitive for the average person. This is one of the reasons so many people are using CoinBase/etc for simple import functionality - this is complicated and borderline scary for people who aren't comfortable with technology. Honestly I never thought it made sense to put this under "Help" or "Debug window" but I stuck with what the new Bitcoin Core was using for consistency. We now know that consistency is great and all... but it can't hurt to go above and beyond. Obviously the average Joe who wants to go import their paper Murraycoin wallet would think those above steps are pretty ridiculous.
I'm still kicking around the possibility of an algorithm change and another fork as well... which of course are both things we will want to do as soon as possible. My concern is simply ensuring the blockchain is secure and somewhat consistent speed-wise - right now it's just being treated as another hit-and-run Scrypt-N coin which is why it's struggling on the exchange - not a lot of exciting news getting crypto people buying on the exchanges... but still profit to be made here-and-there form the hit-and-run crowd. My concerns with merge-mining/etc are that we are then tied to another currency which, despite the advantages of consistency, I'm not sure it makes sense for Murraycoin. I could easily rebuild some new QT wallets with the simple import feature and a modified look... but it would be cool to kill a few birds with one stone. This is something we will have to continue to discuss.