Hello Bractals and Vladimir. I put this problem on the site sometime in October of 2002.
The solution in case the point is on the side of triangle is trivial.
There must be 3 solution lines when the point is "deep" inside the triangle.
When the point is outside or on the perimeter of the triangle where is one solution.
So there must be a locus of points with 2 solutions.
And the construction has to include an angle bisectors.
Bractal, if you use the Sketchpad you can move around vertices
A,B,C and the given point P and see if the arears of the resulting
halves remain equal ( i doubt it is the case for the solution you
provided ).
Anyway, the proof still remains to be seen.
Golland.