Why is viscoelastic litosphere expected to behave as an elastic one in Nakada's Paradox? While reading about True Polar Wander I discovered the so-called Nakada paradox, that states that on 1 Myr time scale the results obtained considering a viscoelastic litosphere don't converge to the result of an elastic litosphere.
If I imagine an immediate elastic response and a viscoelastic response that happens in much longer times (longer relaxation times), I can't figure out why is the viscoelastic litosphere expected to behave as an elastic one and not the contrary (I know that TPW can be forced by various kind of forcings, such as ice melting, and I know that the response of the planet to such forcings is related to Love numbers: but as far as I know the fluid limit of Love numbers is reached asymptotically for great times). 
So, my question is: why do I expect that an elastic litosphere would behave as an elastic one on million year scale if its relaxation time is much greater than the elastic case? 
Some more info: I read some articles, such as "The rotational stability of an ice-age earth" (Mitrovica et al., 2005) and "Ice age True Polar Wander in a compressible and non-hydrostatic Earth" (Cambiotti et al., 2010), but didn't came out with an answer.
 A: I found the answer by myself: I was not considering how high was the viscosity of viscoelastic litosphere in the case of Nakada's Paradox.
While the traditional orders of magnitude of the viscosity for the upper mantle and the lower mantle are respectively about $\nu_{UM}=10^{21} Pa\cdot s$ and $\nu_{LM}=10^{22} Pa\cdot s$, the density of the high viscous viscoelastic litosphere considered in Nakada's Paradox is $\nu_{LS}=10^{26} Pa\cdot s$.
This means that $1\ Myr$ is far below the litospheric Maxwell time for this type of rheology (in my question I wrongly assumed that, at $1\ Myr$ we already reached the asymptotic limit), which is about $30\ Myr$, and that's why we would expect the same behaviour for True Polar Wander considering a high viscous viscoelastic litosphere and an elastic one.
In addition to this, we can compute both the convolution between the elastic tidal Love number with the Heaviside function $K_E^T$ and the viscoelastic tidal Love number with the Heaviside function $K_V^T$:
$K^T(t)=k_2^T(t)\star H(t)$
According to the book "Global dynamics of the Earth", $K_E^T$ and $K_V^T$ coincide not only on the million year scale, but up to $10\ Myr$.
