the 3 images are normal, the script has saved the rejection maps for you so you can check them out. as the dialog box says, it's probably better to take the c_r (or c_d_r) images from the BPP script and feed them into ImageIntegration separately so you can experiment with the proper pixel rejection and normalization settings.
are you clicking the square button on the LRGB process or the round one? you may be trying to apply the LRGB process to an existing image rather than asking it to create a new image.
there are many ways to skin this cat - you can make an RGB image using pixelmath. uncheck "single expression", check "new image" and "RGB color", then put the names of your 3 images into the R/K, G and B fields of pixelmath and execute it. you'll get a new, RGB image.
usually the LRGB tool is used toward the end of processing - you process your L image almost to the end (stretch, sharpen, etc) and also process your RGB image separately. then you put only the name of the L image into the LRGB process and drag the triangle to the RGB image. that replaces the "implicit" luminance in the RGB image with your L image.