The noise function is returning a single "frequency" of noise. The effect on the right (B) is by adding together noise of multiple frequencies, in a "red noise" or "fractal brownian noise" pattern. Low frequencies will be more prominent than high frequencies.
You can change the frequency by multiplying the input to the noise function by a constant. For example,
noise(2*x,2*y) will have frequency 2. So try adding
16*noise1(x,y) + 8*noise2(2*x,2*y) + 4*noise3(4*x,4*y) + 2*noise4(8*x,8*y) + noise5(16*x,16*y) where
new Perlin(…) with different seeds, and see if that's closer to what you want. I don't know for sure whether these constants will be what you need but you can start playing with them to get different effects.