I agree with what others have said, that this is of dubious utility. At best it seems too niche but I'll leave discussion of the tactics to those better versed. I'll rather address some of the engineering practicalities instead.
With an A6M2-N as the base aircraft, assuming no increases in weight from the modified float for fuel carriage, pumps and no structural reinforcement necessary from the increased weight, the 'Rufe' has a difference between gross weight and MTOW of 926lbs. That makes an allowance for only 154 US Gals, only 69 US Gallons or so above that already available from a standard A6M2 drop tank for a great penalty in drag.
E6B, NavLog Calculator, Weather Reports, METAR, TAF, Wind Components, Instrument Simulator, Weight and Balance, Pressure Altitude, Density Altitude, True Air Speed, and a lot more.
e6bx.com
I get that increased fuel isn't the sole object and that operational flexibility is also the goal but still, that is surprisingly low. An earlier effort towards something like the A6M7s wing might be more useful.
You can, of course, up-engine and/or increase wing area for increased MTOW but neither will help your fuel consumption and early-war Japanese engines were quite power-limited.
Let's talk failure modes. Amateurs talk tactics, professionals talk failure modes. If that's not axiomatic, it should be.
When operating as a floatplane, even if temporarily, the main float and pylons must be considered to be primary structural members. The pylons must bear the full weight of the aircraft. They must withstand the drag from the float. That is a lot of wetted area, across two mediums, one of which is much denser and undulating. They will get lateral loading and compression during the take-off run as well but the killer, as with anything to do with aircraft, is always some combination of weight and drag.
So with this concept, you want to make multiple attachment points that
must possess structural strength
frangible?! You are purposely building-in structural failure modes. Frangible attachment points are guaranteed to fail before the aircraft gets “on the step”. The bolts will shear, as they must be designed to do. The forces involved are too great for otherwise, even in the most idealized flat calm.
Use some form of shaped charge? Any charge large enough to ensure a clean separation (of a non-frangible pylon) would undoubtedly cause collateral damage. On an aircraft with the Zero's/Zeke's/Rufe's reputation for flammability?
This really bears re-emphasis. The aircraft's landing gear must be an integral component, stressed to withstand impacts up to the aircraft's structural limits. It's severance can only be achieved by exceeding those limits. How then, is this purposeful vandalism contained, localized and prevented from becoming a loss-of-vehicle event? You cannot design a landing gear to both break and not break. You can have load-bearing or you can have separable. Not both. They are mutually exclusive concepts.
Speaking of separation, all of the drop tank jettisons I'm aware of involve some sort of pitch event. A notional fuel-float must have some residual fuel aboard. On separation, the float will decelerate, the fuel will slosh forward and the nose will pitch down. Except on a float, the forward section is a planing surface. Will this act as a lifting body and pitch the nose up? With such minimal prop clearance? Suffice to say having such a
long object pitching
unpredictably in close proximity to an aircraft before it has time to fall away gives me pause.
What if the notional floatplane is bounced in transit? What's the procedure? Jink and jettison? Separate and then hopefully still be around to jink? What way will it all go, punched off while in a hard bank with a boot full of rudder in?
I haven't mentioned the wing floats and asymmetric hang-ups. Aerodynamicists hate asymmetry.
What may be a theoretical boon seems like a real-world liability to me. It's an interesting concept that falls down the instant you give it to Nakajima to actualise.