Notes. From left to right in the top table, the columns show (i) ξ, a proxy for the evolutionary model (viscous if ξ = 0, hybrid or wind-driven otherwise); (ii) μ₀, the slope of the t_acc,0 − M_* correlation and its implication on the correlation itself; (iii) ζ₀, the slope of the R_d − M_* correlation, and its implication on the correlation itself; (iv) δ₀ = λ_m,0 − λ_acc,0, the difference between the initial slopes of the M_d − M_* and $\dot{M}-M_{\star }$ correlations and its implication on their relative steepness. The final column summarises whether the signs of μ₀ and δ₀ are ‘necessarily the same’ (=), ‘necessarily opposite’ (≠) or ‘can be either’ (○). When discussing the implications on correlations, up(down)wards arrows represent an in(de)crease of the parameters, while horizontal arrows describes the lack of correlation. The different cell colours are purely meant to guide the eye. The top table links the initial conditions, while the bottom table summarises the implications of the evolved difference in the slopes. ^(α)ζ₀ξ is positive, therefore δ₀ can either be negative (implying (a) shallower than (b)) or positive (implying (a) steeper than (b)). ^(b)ζ₀ξ is negative, therefore the same argument as in a holds.