Hey there @MikyLu,
The pressure-velocity coupling is done by the SIMPLE algorithm, which is the only algorithm available for steady-state RANS turbulence models. It doesn’t have an order of accuracy (like the discretization schemes do).
You can change the discretization schemes used for your simulation in Numerics, which can be found in the simulation tree. You can choose 1st/2nd order upwind schemes for the divergence term.
Currently I see that you’re using a full set of 2nd order schemes, which is good for accuracy.
About the initial condition for velocity, I would not change any values there as it sometimes causes instability in the beginning of the simulation with huge variations on parameters. So I usually just stick to using velocity inlet BCs.