How to go about this?

The language doesn't matter ;) i prefer C# but nearly everything can be done in both languages.

1. Input.GetMouseButtonDown(1) is what you need to detect the right mouse button (0==left; 1==right 2== middle). To lock or unlock the cursor use Screen.lockCursor

2. 3DObjects don't have events like mouseover, but you can use Physics.Raycast () in combination with Camera.main.ScreenPointToRay (Input.mousePosition); to get the object you're pointing at. I guess you use the FPSWalker script (or FPSController or how it's names now ;) ). To prevent moving you have to disable the script. Use GetComponent() to access your script and disable it.

3. Here you have to be a bit more precice what kind of GUI you need, how many different objects/GUIs you have/need.

4. Inside one of your GUI.Button() just repeat what i mentioned in 2. Use GetComponent() to access your script and enable it. And lock the cursor again.

*Edit*14.01@03:29

Well if you want a complete different GUI for each object i would use a strategy pattern to offer a common interface. That's quite advanced stuff for a beginner but a basic "tool" in oop. ;) I'm not that deep into Javascript so it's difficult to show you an example in JS but i try:

That's a simple "base class". Every Object have to be derived from this class in order to be able to get selected:

class SelectableObject extends MonoBehaviour
{
   // static var to hold the current selected object
   public static var m_CurrentSelected : SelectableObject = null;
   // define common vars or functions here
   public var Health : float = 100.0f;
}

Here is an example of one of your object scripts:

class Object1 extends SelectableObject
{
   function Update (){
      //Do normal stuff related to your object
   }
   function OnGUI(){
      // check if we are selected, if not, return
      if (m_CurrentSelected != this)
         return;
      // Draw Your GUI for this object here
      if (GUILayout.Button("deselect")) {
         SelectableObject.m_CurrentSelected = null;
      }
   }
}

with this little "trick" you can select an object by it's script component.

Just assign the object you hit with your ray to SelectableObject.m_CurrentSelected

That function is your "picker" script and should be attached to maincamera

function Update() {
   if (Input.GetMouseButtonDown(1)) {
      var ray = Camera.main.ScreenPointToRay (Input.mousePosition);
      var hit : RaycastHit;
      if (Physics.Raycast (ray, hit, 100)) {
         var obj = hit.collider.GetComponent(SelectableObject);
         if (obj != null) {
            SelectableObject.m_CurrentSelected = obj;
         }
      }
   }
}

One last important hint: the script filename and the class name have to be the same. In C# you always have to do that, in JS you don't need to define the class but in that case we want a different base class (in Unity classes have to be derived from MonoBehaviour to be able to attach it to a GameObject)

Good luck! It's not easy stuff for beginners, but everyone have to start somewhere ;)