In normal running, if both power cars are contributing equally to the traction, then if you work it out, couplings near the front are in tension, those near the back are in compression, and those in the middle are neutral.
If the brakes are applied and it takes time for brake effort to spread from front to back, this will no longer apply.
You are not wrong, but only the front power car will be subject to significant aerodynamic resistance which could mean that most of the couplers will actually be under compression from the rear power car.
To get back to the subject, the real issue is the kinetic energy. If e.g. a 225 hits a car on a crossing, it will matter little if the DVT or the 91 are leading. If however, e.g the train derails, hits a concrete bridge support which does not collapse then the way the kinetic energy of individual vehichles is absorbed/redirected will become relevant, such as the way they then derail, which may be sideways and not an end-on compression.
What is of essence is that this has been taken into account in DVT design since the initial Scotrail ballasted Mk2s. The design requirements are to meet safety criteria when in both DVT-leading and power-car leading modes.
The design requirements for other safety systems like signals and junctions are to eliminate derailments and crashes.
If this is a real problem for you then I suggest that when you are in a train with DVT leading you aim for about coach 5 or 6. My logic is that in the concrete block case there is a reasonable chance that kinetic energy will have resulted in some jack-knifing and forces on your coach will be totally survivable. I am not saying that they will not be in other coaches.
Have you ever seen the result of a car crash with a closing speed of e.g 100mph? Probably not as you wouldn't even think about asking the question you did.