Been out a couple of times recently after a hiatus and noticed the meridian flip goes well up until the point of solving and can never find where it is.

That's when I stop things, download an image and solve it manually, only to find out the dec is out by around 20 degrees. RA is perfectly fine

I'm using a HEQ5 Pro running Stellarmate on an RPi 4, Kstars 3.6.0 stable (Thought it would with the package updater)

My current routine is simply: Polar align, reset to home position, slew to meridian to calibrate guiding, solve, slew to target and solve.

Do I need to have a better mount model perhaps and slew to targets further away to build it before slewing to intended target?

Edit: On a side note, I'm not sure why KStars didn't get updated with the package updater + I've checked on the terminal and there's no updates.

Are you running on Astroberry or StellarMate? If it is the latter, you can download and flash the latest version (www.stellarmate.com/support/faq.html?view=faq&faqid=43)

20 degrees is a LOT. The only thing immediately coming up as explanation would be a too tight declination gear meshing that makes the motor lose steps when slewing. Are you close to the mount? Can you hear 'strange' noise?

Using an NEQ6 and can confirm:
Too tight RA/DEC gear meshing makes the motor lose steps when slewing, even without 'strange' noise.

Try loosening the RA/DEC worms so you can feel some slack. And then perform another MF to determine the deviation.
If that works, adjust the RA/DEC worms in small steps to minimise backlash.

It will take some time, but the good thing is that you can do this during the day.