Feasibility of semi-long distance 74HC595 control

I currently have a fully-functioning prototype of a 4-channel RGB flood bank for stage lighting, running through a DB25 cable. I’d like to minimize the wiring and think I could do so by putting the driver PCB (schematic below) on the receiving end of a CAT5 cable, instead of going analog as it is now and sending 13 pins of RGB control (w/one white pin going to a separate dedicated white strip on the flood board) alongside the 12v+ line, necessitating a minimum of 14 wires.

Each flood bank board consists of 4 channels/lamps (198 RGB segments total), each comprised of 9 RGB LED strips (6 segments wide) and 8 white LED strips, in alternating rows:

RGB
W
RGB
W
[etc…]

  • I intend, via the schematic below, to use CAT5 instead of DB25 and relocate the driver PCB (shift registers and transistors) on the back of the flood bank board itself and simply run 5v+, the data/latch/clock lines, 12v+ on single pins and use the remaining 3 conductors of the CAT5 cable for GND. Each board of 4 lamps should consume a maximum of 12~13A (controlled by PWM, which I’ve read makes a full-brightness current load draw around 1/2 that amperage, so ~6.5A), according to my calculations.

[The Big Question]
From what I can tell, since the CAT5 cable will be no longer than 25~50ft, voltage drop should not be an issue, ~7A is within safe limits for around 25’ of 24AWG CAT5 wiring and the consistent timing of the two 74HC595 shift registers should not be a problem at this distance from the microcontroller at the other end of said cable. Is this safe to proceed upon, or should I stick to the DB25 solution? It is worth noting that these boards will be in use for 2~4 hours at a time (fading colors, strobe light functionality and constant flood).

Any input is greatly appreciated… Thank you!