I see no 1.8V source in either schematic, neither of which look
like they are "live" CAD simulation schematics.
If you use uniform W/L then you will not see things work right.
For example the "bus hold" double inverter in the upper pic
will never be overcome by the solo PMOS (vs stronger BH
NMOS) or the double stack NMOS (vs stronger single PMOS
in inverter). So the information on that "X" node may have
only "contention" and "failed" as its logic states.
Similarly half-latch-TINV-based DFFs require that the
feedback be weaker than the forward path to get decent
setup time (or even plain functionality). NAND / NOR
based DFFs do not have this issue, only slowness and
bloat.
"One Size Does Not Fit All" is, I think, today's lesson. But
do not neglect the 1.8V supply's explicit voltage source.