It is important to differentiate between helicity and chirality. Helicity is the spin angular momentum of a particle projected onto its direction of motion. For a massive particle this quantity is frame dependent. Furthermore, since angular momentum is conserved, as a particle propagates helicity is conserved.
On the other hand, chirality is an innate property of a particle and doesn't change with frame. However, the mass term for a Dirac particle is,
\begin{equation}
-m(\psi_L^\dagger \psi_R + \psi_R^\dagger\psi_L)
\end{equation}
(in this notation the Dirac spinor is $\Psi= (\psi_L , \psi_R) ^T $). This term can be thought of as an interaction term in the Lagrangian which switches the chirality of a particle (e.g. a left chiral particle can spontaneously turn into a right chiral particle)
For a massless particle, chirality is equal to helicity.
With that background we can finally address your questions.
- Both helicity and chirality definitely make sense for a massive
Dirac spinor. However, that doesn't mean that a Dirac spinor is a
helicity and chirality eigenstate. In the same sense that energy
makes sense for a particle, but it may not be an energy eigenstate.
- As you mention the left chiral and right chiral fields can't be
decoupled from each other due to the mass term. The mass term can
always switch a right handed field to a left handed field and vice
versa.
- As I said above, the helicity of an electron is indeed frame
dependent. So it may look like a left or right helicity electron
depending on the frame, however its chirality is not frame
dependent.
- If we write the Dirac Lagrangian in terms of chirality eignestates
then we have, \begin{equation} {\cal L} _D = i \psi _L ^\dagger
\sigma ^\mu \partial _\mu \psi _L + i \psi _R ^\dagger \bar{\sigma}
^\mu \partial _\mu \psi _R - m \psi _L ^\dagger \psi _R - m \psi _R
^\dagger \psi _L \end{equation} Then we can think of $\psi_L$ (left
chiral particle) and $\psi_R$ (right chiral particle) as two
different particles that can turn into each other spontaneously
through a mass term. Putting them together, into a Dirac spinor
masks this property. However, they are still well defined
separately.