In the "Create Objects" interface, check the "barycenter" option under Reference Object.
 
If you don't do this, then when Gravity Simulator converts your orbital elements to state vectors, it uses the mass and position of the Reference object only.  With barycenter checked, it adds the mass of the star + the white dwarf, computes their center of mass and uses that for the position.
 
Like I said in another thread, Gravity Simulator does not yet provide output for barycentric systems.  So your results will be orbital elements with respect to a single object, rather than with respect to a barycentric system.  (i.e. they'll be meaningless unless you're really good with Excel and the conversion between orbital elements and state vectors, as the math can be done externally to the simulation).
 
It's not that hard for me to incorporate such output as it would use existing blocks of code.  Just haven't gotten around to it yet.

It's much easier to do the conversions in a programming language than in Excel.  I'll get some coding time after the school semester is over.
 
You can still draw some qualitative conclusions from the screen plots.

Imagine a barycenter as a ficticious object.  It's mass is the combined mass of 2 objects.  It's location is the center of mass of those 2 objects.  The center of mass is just the weighted average of the positions.
 
In reality a barycenter is not an object.  It is just an empty point in space.  But any object that orbits the pair of objects from a great enough distance will behave as if it is orbiting the ficticious object.
 
For example, let's create a 5 object system:  
Star A: component 1 of the binary system.  1 Msol
Star B: component 2 of the binary system. 0.9 Msol in a tight 1AU binary orbit.
Planet Aa: Orbits star A
Planet Ba: Orbits star B
Planat ABa: Orbits the pair of stars from a distance.
 
File > New
 Creates a 1 solar-mass star
 
Objects > Edit Objects
 Change name to Star A and change color if you like
 
Object > Create Objects
 Name: Star B
 Color : your choice
 Reference Object: Star A, leave "Barycenter" unselected
 Mass: 0.9 Msol
 SMA: 1 AU, a close binary system
 Ecc: 0.3, moderately eccentric
 Inc: 0, best to define the plane of the star orbits as the ecliptic
 Size:1000000 km, this is not important
 All others: not important
 
Press Create
 
This creates the binary pair.  Gravity Simulator uses their combined mass when converting the orbital elements you provided into the state vectors that Gravity Simulator uses to crunch the numbers.
 
Objects > Create Objects
 Reference Object: Star A, leave "barycenter" unchecked
 Name: Planet Aa
 Color : your choice
 Mass : 1 Mearth, or your choice
 SMA  : 0.1 AU, or your choice, but keep it close to the star.
 Ecc   : 0, or your choice, but make sure that its apastron doesn't wander too far from the star.
 Inc    : 0, or your choice, but keep it low so Kozai doesn't destabalize it
 Size   : 12000 km, or your choice
 All others: unimportant
 
Press Create
 This creates a planet in orbit around star A.  Gravity Simulator uses the combined masses of Star A and Planet Aa when converting the orbital elements you provided into the state vectors that Gravity Simulator uses to crunch the numbers.
 
 
Objects > Create Objects
 Reference Object: Star B, leave "barycenter" unchecked
 Name: Planet Ba
 Color : your choice
 Mass : 1 Mearth, or your choice
 SMA  : 0.1 AU, or your choice, but keep it close to the star.
 Ecc   : 0, or your choice, but make sure that its apastron doesn't wander too far from the star.
 Inc    : 0, or your choice, but keep it low so Kozai doesn't destabalize it
 Size   : 12000 km, or your choice
 All others: unimportant
 
Press Create
 This creates a planet in orbit around star B.  Gravity Simulator uses the combined masses of Star B and Planet Ba when converting the orbital elements you provided into the state vectors that Gravity Simulator uses to crunch the numbers.
 
 
Objects > Create Objects
 Reference Object: Star A, this time select "barycenter".
 Name: Planet ABa
 Color : your choice
 Mass : 1 Mearth, or your choice
 SMA  : 10 AU, or your choice, but keep it far.
 Ecc   : 0, or your choice, but make sure that its apastron doesn't wander too far from the star.
 Inc    : 0, or your choice.
 Size   : 12000 km, or your choice
 All others: unimportant
 
Press Create
 This creates a planet in orbit around the barycenter of star A and star B.  Gravity Simulator creates a ficticious object whose mass is the combined mass of Star A and Star B, and whose position is the weighted average of the position of Star A and Star B.  It then uses the combined masses of the ficticous object and Planet ABa when converting the orbital elements you provided into the state vectors that Gravity Simulator uses to crunch the numbers.  The ficticious object is not used in the number crunching, only in creating the object, for the purpose of computing the state vectors of the planet.
 
You should now have the 5-object system described above.
 
When viewing the system, select Star A as the focus object.  Then use the F/A button on the Graphics Options interface to toggle back and forth between Floating Mode and absolute Mode.  Pressing F places you in the floating mode, and the screen is centered on the system barycenter.  Pressing A returns you to absolute mode and the screen is centered on Star A.
 
 
 
In the case of your system, all the planets orbit the pair, so you would select barycenter when creating all objects except the white dwarf.

Huh. I did what you said, and the planets all immeditaely shot off at tangents to their orbits with a 0.25 second timestep...!  
 
Here's the new gsim file, can anyone identify what's wrong?  
regina2.gsim

*scratches head*
 
Thanks for posting all that, but that's almost exactly what I did. The only differences are that Lusor was 1.51 solar radii, I actually entered "0" into the Eccentricity and Inclination boxes for all the objects, and I put a leading "0" before all the values that were less than 1 (i.e. 0.626 instead of .626). Otherwise it was the same.  
 
I'll try to make it again following your instructions exactly and see if it still screws up.
 
Though as a suggestion, one thing you might consider is to actually make the barycentre an object in the program - that way we can explicitly tell it to use it as the reference object instead of the star.