I ran the numbers for an enclosed Dyson Sphere with the same specs as the Stellar Engine. The primary difference between an enclosed Dyson Sphere and the Stellar Engine at present is that, for the enclosed case, heat radiated off from the receiving face is absorbed by another part of the Dyson sphere, rather than lost into space. This means it operates at a much higher temperature. In fact, that temperature happens to be slightly hotter than the surface of my star (although that is a bit of a coincidence), at about 4200K, which is twice as hot as the current temperature. Because of this, it produces over 5 times more power per unit area, although melting may be a concern. Conveniently, increasing the distance between the sphere and the star reduces the operating temperature, so the operating temperature can be calibrated like that, trading off with efficiency.