In fact, the conductivity is related to the dielectric permittivity as
σ=ω ε0 ε*(1-1/tgδ), i.e. the conductivity is included into the imaginary part of complex dielectric permittivity: ε*=ε0[ε'-i(ε"-σ/ω)].
Therefore, you can make the transition from conductivity to dielectric constant (ε') only at very low frequencies, when Debye dielectric loss ε" may be neglected, i.e. ε*=ε0(ε'-iσ/ω)
In this case, if you know tgδ at very low frequencies, roughly ε'=σ/(ωtgδ). Problem is that doing such a thing makes no sense because it will be low-frequency dielectric permittivity, which is about the static (at zero frequency) dielectric permittivity that you have to measure anyway to define the tgδ. Not to mention, ε(0) would give you a little at higher frequencies when you have to know ε(ω).
probably the answer is much simpler - you're talking about the specific admittance (reverse specific impedance), Ys, which you can measure. For the dielectric, one may consider Ys* = iωCs* = iωε0(ε'-iε"), i.e. ε' corresponds to the imaginary part of Ys*, Im(Ys*) = ωε0ε' =ωε0|Ys|/(1+tgδ), where Cs* is the specific complex capacitance and |Ys*| is the magnitude of Ys*. Looks like you are confusing yourself with the conductance, σ, which you can measure only at zero frequency (DC).
Microscopically yes, Aseel, you understood this right.
Macroscopically the conductivity leads to the so-called interfacial (Maxwell-Wagner) polarization or so-called absorption polarization (space charge at electrodes) which affect the measurement, mostly the imaginary part of it.