please have a look at the link added in post #36 to see how the noise/measurement bandwidth affects the RMS noise. It is important to set the appropriate bandwidth (start & stop frequency) in LTSpice while performing a noise analysis. Have also a look at [1] and [2] to see how to perom a noise analysis and how to determine the RMS noise.
100mV on the initial circuit with 4 OpAmp with G = 10 yes I remember. However if I reduce the gain on each single stage at least in the first stages the noise should be lower. And it is for this reason that I redesigned the stages with G = 2. Obviously, with the same OpAmp, the gain will be greater but maybe I just need less.