MBP5,2 updated to 10.14.5b5 from b4. All fine so far.
Here are some KP observed on MBP5,2 with 10.14.5b4 and b5. Both betas really very similar. Both are stable-enough for me when using only the 9600M GT GPU and avoiding Safari.
1) in 9600M GT mode total of 17 more KPs since 3 May, i.e. after I had erased the SSD and re-installed from scratch. Used the 10.14.3/1.2.3 dosdude USB installer, then updated to 10.14.5b4 OTA and post-install with 1.3.1 (with force rebuild cache), on 7 May updated to b5 same way.
KPs seem to correlate with Safari usage, although none of them shows it as process name. Sometimes Safari would run well for hours without KP, often shorter, and KPs happened after reboot even without using Safari again. Did a defaults delete com.apple.Safari then which (I think) improved stability again.
If not kernel, process names are 3* corespotlightd, 3* system_profiler, sysmond or clamd (antivirus).
I normally used also Sophos antivirus, and as the recently loaded kext was 4* sophos, I de-installed it; still KPs after.
2) switched once to 9400M GPU (system settings, logout/login -> gfxCardStatus shows still d for 9600M GT, shutdown and boot -> i for 9400M now.) KP few minutes after boot, with GeForceTesla kext in backtrace, report is very short and attached. That was with .5b5.
kexts in backtrace were: IOStorageFamily, apfs, AppleMobileFileIntegrity, 4* AGPM, GeForceTesla (with 9400)
kexts last loaded: 7* msdosfs, 4* sophos, 4* AudioAUUC, AppleXsanScheme
kexts last unloaded: 2* msdosfs, 6*AppleUSBMergerNub, AppleXsanScheme.
zip of all KP reports also attached.
In summary, I'm not surprised given Syncretic's post #14936. But it seems that MBP5,2, at least my one, is less stable then other machines with 10.14.5 mentioned in this thread. Note that I did fsck -y and AHT on my machine, showing no errors.
Still hope not to go back to 10.14.3 though. I start using the Hopper disassembler to learn more about my system (I found that a lot easier at Motorola 68k times. I noticed dosdude has member number 68000?).