This pass should be the last. The latest schematic is posted in the project details.

Change List:

1. Replaced the MCP1702 3.3V LDO with the TPS70950 5V LDO from TI. The TI LDO has a lower Iq (1µA), higher VIN (30V), but costs $1.04 vs. $0.51 for the Microchip part. The output  capacitor needed to change since the TI LDO requires between 1.5µF and 47µF for stability. The increase in VDD from 3.3V to 5V is intended to lower the RDSon of M3-M5 for input voltages above 3.3V.
2. The symbol for the D-type flip flop was changed to reflect the function.
3. The layout was altered significantly.
   1. The LED was moved to the bottom side since it should be toward the inside of the cockpit area to get better heat relief, and the hall-effect switch on the top side should be toward the outside of the fuselage to improve sensitivity to the magnet.
   2. M1 and M2 were moved closer the the B1 and B2 pads.
   3. The VOUT pad was located much closer to the sources of M1 and M2.
   4. The GND pads were moved much closer to the sources of M3-M5
   5. M3-M5 were reconfigured into a "U" shape to reduce the number of squares of metal trace to the sources and drains.
   6. I took care to not locate components in the path of the lead wires to the pads, which are now located in the interior of the board. This is a reliability issue.

I'm expecting the switch losses to decrease significantly. The board actually got a tad smaller: 0.481" x 0.869".

The maximum input voltage is now limited to 20V by M1 and M2, but the voltage rating of C6 lowers it even more to only 16V, so I reduced the specification accordingly to 13V -- a 3S LiPo. If the voltage rating of C6 was increased by either paying more or lowering the value to around 10µF then VIN could be raised to 20V.

The PCBs have been ordered in 2oz. copper. They should arrive in a few weeks.