I'm using a LJ12A3-4-Z/BY (PNP with a diode similar to yours), on my printer:
; Endstops
M574 X1 Y1 S0 ; Set active low endstops
M574 Z1 S0 ; Set endstops controlled by probe
M558 P3 H5 F120 T6000 ; Set Z probe type to modulated and the dive height + speeds
G31 P250 X16 Y30 Z0.5 ; Set Z probe trigger value, offset and trigger height

threw the whole config out the window and started all over, worked alot better.
still had tons of wifi disconnects though, strangely they stopped immidiatly when I disconnected the LCD display

dc42, they are unchanged in both firmwares.
but I can post them if you like:
homeall.g:

; homeall.g
; called to home all axes
;
; generated by RepRapFirmware Configuration Tool on Sun Jun 18 2017 15:09:50 GMT+0200 (Vest-Europa (sommertid))

G91 ; Relative positioning
G1 Z5 F6000 ; Lift Z by 5mm
G1 S1 X-305 Y-305 F3000 ; course home X or Y
G1 S1 X-305 ; course home X
G1 S1 Y-305 ; course home Y
G1 X5 Y8 F600 ; move away from the endstops
G1 S1 X-10 ; fine home X
G1 S1 Y-10 ; fine home Y
G1 Z-300 F1800 S1 ; Move Z down until the switch triggers

G90 ; Absolute positioning

; Tell the firmware where we are
G92 Z0.5

; Uncomment the following line to lift the nozzle after probing
;G1 Z1 F100

Homez.g:

; homez.g
; called to home the Z axis
;
; generated by RepRapFirmware Configuration Tool on Sun Jun 18 2017 15:09:50 GMT+0200 (Vest-Europa (sommertid))

; Lift Z relatively to current position
G91
G1 Z5 F6000

; Move Z down until the switch triggers
G1 Z-300 S1 F1800

; Back to absolute positioning
G90

; Tell the firmware where we are
G92 Z0.5

; Uncomment the following lines to lift Z after probing
;G91
;G1 Z1 F100
;G90

Upgraded to the new firmware 1.20 in an attempt to see if the massive AJAX errors I have in 1.19.2 would be cured.
Instead I had a new issue: the Z probe does not stop the homing at all, it did in 1.19.2 and 1.18.2.
I run a CoreXY setup, and when I reverted back to 1.19.2 the probe works just fine.
in 1.20 I tested the input status via 119 and it reported the probe as triggered but during homing it ignored it completely.

• If you currently have G31 parameters for your active Z probe in config-override.g that are different from the ones in config.g, you should copy them to config.g, otherwise they will be lost next time you run M500.

I do not have a config-override.g file, but apart from that I don't quite see why it should completely ignore the probe…
a copy of the settings in 1.19.2:
; General preferences
M111 S0 ; Debugging off
G21 ; Work in millimetres
G90 ; Send absolute coordinates…
M83 ; ...but relative extruder moves
M555 P2 ; Set firmware compatibility to look like Marlin

M667 S1 ; Select CoreXY mode
M208 X0 Y0 Z0 S1 ; Set axis minima
M208 X300 Y300 Z295 S0 ; Set axis maxima

; Endstops
M574 X1 Y1 Z0 S0 ; Define active low and unused microswitches
M558 P3 X0 Y0 Z1 H5 F120 T6000 ; Set Z probe type to switch, the axes for which it is used and the probe + travel speeds
G31 P250 X16 Y30 Z0.5 ; Set Z probe trigger value, offset and trigger height
M557 X15:285 Y15:285 S20 ; Define mesh grid

I'm using a CoreXY, so modified mine to match the https://duet3d.com/wiki/Configuring_RepRapFirmware_for_a_CoreXY_printer
like I said, all the homing files are directly from the Reprap configurator.

basicly modified mine to:
G91 ; Relative positioning
G1 Z5 F6000 ; Lift Z by 5mm
G1 S1 X-305 Y-305 F3000 ; course home X or Y
G1 S1 X-305 ; course home X
G1 S1 Y-305 ; course home Y
G1 X5 Y8 F600 ; move away from the endstops
G1 S1 X-10 ; fine home X
G1 S1 Y-10 ; fine home Y
G1 Z-300 F1800 S1 ; Move Z down until the switch triggers

G90 ; Absolute positioning

this solved it yeah, dunno why the reprap code generator did it this way, as the other ones are that way aswell, eg:
; homex.g
; called to home the X axis
;
; generated by RepRapFirmware Configuration Tool on Sun Jun 18 2017 15:09:50 GMT+0200 (Vest-Europa (sommertid))

; Lift Z relative to current position
G91
G1 Z5 F6000
G90

; Move quickly to X axis endstop and stop there (first pass)
G1 X-305 F1800 S1

; Go back a few mm
G91
G1 X5 F6000
G90

; Move slowly to X axis endstop once more (second pass)
G1 X-305 F360 S1

; Lower Z again
G91
G1 Z-5 F6000
G90

completely ignores it in coarse, I can manualy actuate the end stop and see the diode light up on both the end stop and on the duet.
as for fine home I dunno since it fails to coarse home

Ok, so for some strange reason this happens:
It starts the correct sequence, starts to lower the Z height by 5mm, then starts to move X and Y towards -305 as the homeall.g file says.
But for some reason it completely ignores the Y end stop. (X works fine, as does Z)

Fine I thought, could be a fault in the switch, however M119 says triggered.
also running the home Y individualy works fine, aswell does all the other axis.

is there something I've missed here?
a direct copy of the homeall.g file (generated by RepRapFirmware Configuration Tool, so not modified by me) :
; Relative positioning
G91

; Lift Z
G1 Z5 F6000

; Course home X or Y
G1 X-305 Y-305 F1800 S1
; Course home X
G1 X-305
; Course home Y
G1 Y-305

: Move away from the endstops
G1 X5 Y5 F6000

; Fine home X
G1 X-305 F360 S1
; Fine home Y
G1 Y-305 S1

; Move Z down until the switch triggers
G1 Z-300 F1800 S1

; Absolute positioning
G90