I am currently researching the feasibility of setting up a coilgun that launches Cubesats for a paper. 1 cubesat per launch total payload mass <100kg.
It is certainly possible to mount electronic components to survive 100G, see HARP which had instrumented rounds firing at thousands of G. The fundamental step is to mount everything in epoxy with the aim to have everything be roughly the same density with no air gaps or denser sections to keep forces in direction of motion, not attempting to displace things sideways which tends to break electrical connections.
This can work for a single function sensor but some interesting complications making a useful cubesat fly this way. Optical sensors for example generally need a void for the lens assembly to work, so do many solid state MEMS devices and crystal oscillators . Power will also be complicated, since most high density battery chemistries are high density by virtue of mechanically minimalist design and loose a lot of capacity being made robust enough to not crumple to one end of the cell and short out. High performance solar cells are also not noted for robustness and would need to go on the outside of the chassis, where they would tend to sag away as the chassis shortens during acceleration.
So a 100G target is provably possible, but the question is how useful the resulting electronics is at actually being a functional satellite, especially with probably a much lower electric power collection and storage budget compared to a rocket lifted cubesat of the same size.