I watched Keep Making's (Multiboard creator) video about multi-material stacked printing.
Here is the corrected version:
I have spent some time optimizing this process for other models and offer this optimized, albeit considerably more complicated, approach. I've done the hard work, so you don't have to.
I am using a BambuLab X1 Carbon printer with an AMS unit. Standard printing speeds are in excess of 200mm/s, which is too fast to get PLA and PETG to stick together reliably.
Keep Making addressed this issue by over-extruding the PETG layers. That may work, but there's a better method.
The key to getting PLA and PETG to stick together well, but not too much, is to adjust printing speeds, treating the first layer of each material as if it were the first layer on a build plate. PETG atop PLA is performed at 80mm/s, and PLA atop PETG at 60mm/s. PETG prints hotter than PLA, hence the faster speed to avoid sticking too much.
Getting this to happen in the slicer requires multiple modifier "sheets," which are used to reduce the speed of the first layer of each tile in the stack. The PETG layers are a separate part and simply have their printing speed reduced to 80mm/s for all features.
Further Print Time Optimization In a later version of a Core tile stack, I changed the PETG print speed back to default (200mm/s+) and added four more 0.2mm heigh modifier parts to reduce the speed of only the first PETG-atop-PLA layer, rather than the whole two layer separator part. For a full 8x8, this saves a bit more print time.
Keeping the prime tower smaller also saves filament and print time. A width of 60mm is plenty.
FLUSHING VOLUMES You should have these set to around 580 for both PETG-to-PLA and PLA-to-PETG. By default, lighter to darker color changes are much less. Too little purge results in a mixture of PETG and PLA for the first part of printing the first new layer of the new material, which in turn can result in undesirable layer bonding.
A tidier alternative here is to reduce the flushing volumes and use a larger prime tower or increase the priming volume significantly. (Tidier only in that you get less "poop" but still use the same amount of purge filament in the prime tower instead.)
Other Comments This method contains only eight (8) filament changes for the entire stack. The printing profile is the Keep Making recommended 3 walls and 15% infill.
I've included the separator object as a separate file in case you want it for your own experiments, but it is NOT part of the model supplied here for printing. Just load the .3MF.
For a smaller test print, simply use your slicer to "cut" the model a couple of times and discard unwanted larger pieces.
The large black sheets between layers as shown in the 3D viewer here at Thangs are actually speed modifier parts and not part of the actual printed geometry.