The discussion here is one of fluid dynamics. Basically, the fluid flowing through a pipe, or the surface of an airplane, is not uniform. There are boundary layer effects. Fluid (air in this case) is nearly static within the boundary layer. This is bad for a turbofan because if it doesn’t get uniform flow then part of it doesn’t work as efficiently, resulting in poor fuel efficiency and excess wear. Everyone tries to get around this in different ways. The F-35 uses a DSI because it is a cheap means of reducing RCS. The circular bulge simply forces the boundary layer on the airplane’s skin to flow more in uniform with the air outside of the boundary layer. This effect rapidly starts to deteriorate as you approach Mach 2. So the J-20 and the F-35 are likely bound by that no matter how powerful of an engine/engine’s they receive. Other designs like the F-22 and the Su-57 use a diverter. This is basically a gap which separates the inlet of the aircraft from the surface so that the boundary layer has no effect. This allows for speeds a little beyond Mach 2, but at a cost of making it more difficult to maintain stealth at lower frequencies (the F-35 is said to actually be stealthier than the F-22 in lower frequencies).

All of these designs are fixed inlets. To propel a turbofan aircraft up to the velocities where ramjet propulsion is required you need a variable inlet. However, this is practically impossible to implement in a LO or VLO design.