Enabled mesh bed leveling which does help with the bed adhesion. As I have no sensors, I use the manual leveling, which actually works very well, just takes bit more time.
Using Marlin's linear advance, currently testing value K150 (~60cm bowden tube). Does reduce the need for retraction, usually with this long bowden I need to use something like 5mm retraction. With LA enabled, even 2.5mm is enough.
Replaced A4988 stepper drivers with DRV8825 for X and extruder. Does make the printer bit quieter. Will swap Y driver to DRV8825 later, need to test it properly because in past I have had some bad experience with driving to parallel motors with DRV8825.

Some prints that I got done today on the new Dollo:

Stackable box, looks like crap for a couple of reasons: retractions for PETG and linear advance not tuned in, warping and inconsistent bed movement (bed does not stay level in Y-axis due to bent z-screws). I actually printed this the see how much effect the last problem has and it can be seen that on 3mm intervals (z-screw pitch) there's some banding happening. Other than that, z seems to work fine and the dimensions are very good.
https://www.dropbox.com/s/ihoysq5yej...21.34.jpg?dl=0
https://www.dropbox.com/s/ihl0e1l26o...21.44.jpg?dl=0
https://www.dropbox.com/s/6f0emx127p...21.55.jpg?dl=0

Of course 3DBenchy is one test model that just needs to be printed. PETG and 0.2mm layers, S3D. Turned out quite good actually. The z-banding can be seen here too, and again retractions need a bit of tuning. Also no part cooling yet, even though PETG isn't that dependent of it.
https://www.dropbox.com/s/j3duq4flso...39.13.jpg?dl=0
https://www.dropbox.com/s/fy2v3o0jeg...39.23.jpg?dl=0
https://www.dropbox.com/s/bl45m8jeir...39.34.jpg?dl=0
https://www.dropbox.com/s/o53t6btgxk...39.45.jpg?dl=0
https://www.dropbox.com/s/xgmrev0mak...57.40.mp4?dl=0

Printing parts for the cable shroud. Most of the parts were printed on Prusa MK2, but now I can use Dollo to print them. PLA, KISSlicer, 0.25 layers (KISS dynamic layers). Also tuned linear advance from 125 to 150 for these parts.
Also when positioning the models in line with the z-screws, the bed rocking motion is minimal.
https://www.dropbox.com/s/61o1tyvt4y...33.15.jpg?dl=0
https://www.dropbox.com/s/jcy8o5edm2...33.26.jpg?dl=0
https://www.dropbox.com/s/h3zb8qno5h...58.48.jpg?dl=0

To fix the bed rocking motion, I'm printing new z-screws that use flexible coupling to the motors. Also because the z-screw consist of multiple parts, connecting them together is prone to result in bend screws. I designed a jig to keep the parts straight when the glue sets:
https://www.dropbox.com/s/srcf56s4bj...04.19.jpg?dl=0
https://www.dropbox.com/s/qvdz1sxgry...04.34.jpg?dl=0
https://www.dropbox.com/s/6qpp88pce8...04.38.jpg?dl=0
https://www.dropbox.com/s/6rck3jrbsxba5h9/2018-07-05%2012.12.58.jpg?dl=0

Next:
- install new z-screws
- order a heated bed
- mounts for RAMPS and Raspi
- mount for ATX power supply
- design some parts for cable management
- part cooling