I just dont see why the R-P-S dynamic does not extend to big ships too. Instead they are left with an Achilles Heal
The RPS dynamic applies even less to smaller ships. When the Viper was my primary vessel I mounted nothing but fixed beam lasers because fixed beam lasers stripped shields, and could do it despite SCBs. The fact that it took longer to harm the hull was almost completely irrelevant against smaller ships because 'longer' was usually more than fast enough. Only larger ships have enough hull strength to really even bother with selecting weapons for damage vs. hull.
The Anaconda, from certain angles, is quite vulnerable, but most larger ships capable of combat are less so, and a competent Anaconda pilot should not be exposing the power plant to direct attack in the first place. It's very difficult to take out a CMDR in a Clipper, FDL, or Python with subsystem targeting faster than tearing apart the hull.
It's all linked though, because having purposeless bulkheads/reinforcement mean that we don't fight without shields. And since SS are invulnerable while shield is up, the whole SS targeting also becomes almost pointless. Gah.
There is no permutation of mechanisms that would still result in this game being Elite: Dangerous where a focus on shields would not be wise for the overwhelming majority of combat setups.
Sheilds constantly recharge. Hull does not. Shields are a renewable resource. Hull and module damage is not. This, in and of itself, is more than enough to justify making your shields as strong as reasonably possible, and for disengaging before or immediately after shield collapse.
That said, there are still some shieldless setups that can be quite formidable. Can't reliably target subsystems, at least not with auto aiming weapons, on a ship that is constantly falling off sensors because it's pulsing silent running and/or dropping heatsinks. Combined with armor, hull reinforcers, and B class modules in critical areas, there are shieldless ships that can go toe to toe with their still shielded brethren.
What else is there to put in the internal modules?
That's precisely his point.
Though one AFM honestly wouldn't be a bad idea.