I put in a 1W R2H HD into a KL4 bezel driven by a Wizard 400, here are some beamshots (Note that what appears on screen are not accurate representations of actual tints):
Have you done any runtime tests? I'm curious if the Wizard's lower efficiency (despite it's boost once real Vin drops below the R2's Vin) means it poops out sooner than a DB400.