Ignorant troll remarks notwithstanding, this idea is sensible and doable, but requires deep understanding of TCP: e.g. yes it is supposed to be robust to duplicated or misordered packets, but no it will not perform acceptably with even low incidence rates of either. Symmetric PEPs w/explicit multipath packet processing will be needed: the Windows client is not a good candidate for one end of that pair; a lightweight Linux based transparent gateway, using e.g. tun/tap and netfilter, with some coding, can do it. We have been doing similar things for disadvantaged mobile wireless platforms (e.g. aircraft in flight) for years. Beware: transport mode IPSEC, if running between the ultimate client & server, makes it much more difficult (requiring guesstimates of what is happening with opaque transport layer headers). With an IP tunnel over a Type II Hybrid ARQ/FEC transport over packet by packet concurrent multipath routing, you can accomplish the OP's goal & more, but overhead is more than 100% unless you have 3 or more paths.

VeriKrypt would be a translation into mixed Latin & Greek.

I & my company are dedicated to open standards, open source software & open source hardware. We also need to make a living, which has been very difficult for too many years now. Where would you recommend I look for information with which I can educate myself on monetizing technology development while remaining true to my cypherpunk & libertarian philosophy? Regardless of what some (who presumably have never had to make payroll) may say, not all information wants to be free, as it will never be created if its creators know their reward for creating it will be continuing poverty. We also recognize that pure money, useful information & free energy (in the statistical mechanics sense) are equivalent, so it ought to be possible to convert some of the useful information we generate into money...

Multics may be dead but its genetic material lives on. Indeed some of the recessive genes, long dormant, are resurfacing. Android tries to implement capability based security, but as Android is typically layered on top of an operating system (usually Linux) that does not natively support capabilities from the hardware on up, Android can't really enforce capability based mandatory access controls. The L4 microkernels, OTOH, do... and they offer a fine base for virtualization frameworks in which more familiar OSes (and "bare metal" apps) can be hosted. So I chant horrific syllables by moonlight at forgotten altars of the Old Ones...