Brainbow and Flybow are great tools for in vivo analysis of such events because they provide labeling of distinct cell types with a myriad of colors. They have varieties which label neurons and varieties which label astrocytes. The neuronal version is problematic in the sense that each neuron may express one color out of a spectrum of colors (64, I think) making standard fluorescent/antibody labeling and investigation difficult (if not impossible for most investigators do to microscope sensitivity and laser line limitations). The astrocytes are nicer in the respect that each astrocyte may express one of three distinct colors making standard staining and analysis much more tractable. You might be able to isolate SVZ cells from the neuron version to answer your questions but you'd have to REALLY want the answer because you'd likely have to start a breeding program and cross the Brainbow with some Cre mice. In our experience, its a bit of a pain and can be prohibitively expensive. Jax has the mice at http://jaxmice.jax.org/findmice/brainbow.html.
As an alternative you might try to express a gfp reporter with a neuroprogenitor specific promotor. You'd have to do some research on that one. If memory serves, I think ErbB4 might be one-- don't quote me on that.
The hand waving to all this is the assumption that in vitro models, be they immortalized cell lines, stem cells or dissociated primary cells, behave the same way as cells in vivo. Usually this isn't the case but depending on what your questions are and how you employ your madels to answer those questions, you can still get valuble data.
With all that said, its likely the case that the best in vitro system you can generate and maintain is a stem cell derived one. Reference PubMed for resources, contact authors and labs and good luck. FYI, There's a Stem Cell and Regenerative Medicine Institiute in Seattle which has good success with such endeavors-- plus they are always eager collaborators.
Depending on what kind of neurogenesis you wish to look into, you will have to choose an appropriate cell type.
There's a well referenced book available from Springer that has a chapter describing some of the cell lines with its uses called "Protocols for Neural Cell Culture" that I would recommend.
Additionally, you could also use primary cells from the ventricular zones and culture them as neurospheres. I would direct you to a Nature Protocols on this : "Neural Stem Cell Culture: Neurosphere generation, microscopical analysis and cryopreservation."
Finally, quite a few labs have developed protocols to generate cerebral cortical neurons from embryonic stem cells as well as iPS cells. For those, the earliest works are from Eiraku et al. 2008 and Gaspard et al. 2008. Both use in-vitro systems to generate neurons from mouse and human ES cells. So far, I would say these are the best systems to recapitulate what happens in vivo though there are issues with some cell types being absent and also the duration of the protocol.
Depends also on which model you are interest (human, mouse, rat, other?). Mouse p19 cells (treated with retinoic acid) are easy to use and can be used as a model of generic neuronal cell differentiation. Another possibility are the rat PC12 cells.
My old lab dissected SVZ neural stem cells from postnatal or adult mouse brains and cultured those. Other labs study neurogenesis in vitro in the hippocampus- same thing, just dissect the DG neural stem cells and culture those in vitro.