Here is an interesting story of my views on the last 15 years of CG history that may shed a little light on the condition of Carrara.  I want to emphasize that this is strictly my opinion and that I am making these observations from a very limited viewpoint.  However, it might be interesting nonetheless.
I would classify the Computer 3D Artwork (C3DA) community into three general categories (for the sake of this discussion).
1.)  The BIG guys.  These are outfits like Industrial Light and Magic, Weta, Pixar, etc..  I'm not even going to talk about these guys because they're in a whole class of their own which is really outside of the Carrara community in most practical respects.
2.)  The private pro's.  I consider these people a group of people that are proffessional C3DA people but they pretty much don't write their own software.  This group comprises of advertising agencies, consultants, etc..  The vast majority of people like this use many software packages of which Carrara is just one.
3.)  The hobbyists.  That would be people like me who have jobs which are not part of the C3DA pro community but who dig C3DA, have invested a little effort, time, and money in it, and crank out some art here and there.
I will start this story in the late 90's.
Two things happened in the late 90's which had a BIG impact on the C3DA community.  They were,
A.)  Computer power got big enough and cheap enough
that
B.)  Affordable C3DA software packages could produce artwork that was starting to gel.
Earlier than this, computing power that had been used for artwork commonly comprised of large computer systems which were really expensive and C3DA software was also pretty spendy.  When 486 computers hit the street, we started seeing some things change.
New C3DA software really started to bloom like Bryce, Raydream, Terragen, Poser, and Blender (to name some).
I can remember Lightwave costing many thousands of dollars (well above $5,000) with plug-ins for things like fur costing $800 an up.  For an average dude, $5,000 software was NOT in the cards.  However, $300 for Bryce was a whole different story.
In the time frame of about (let's just say) five years, we saw a huge new C3DA community become born.  Before this, people just plain couldn't afford it and computers for home just couldn't take it but it all kind flipped over.
Back then, there were no college courses in C3DA, the number of people who were knowledgeable (I mean seriously knowledgeable) were few in number, and that meant that there was a gob of people who were sort of solo reaching out to the internet for communities of people to learn and get moving.
Websites like Renderosity were super important then because these places were some of the few living and thriving refuges where people could get information.  It was a pretty cool time.
Computer power was really growing on a yearly basis, the C3DA community was also growing a lot and the software vendors followed along.
Then, we saw was an interesting divergence in the C3DA community.
The people who were serious about it started turning C3DA into businesses and proffesions.  They became group #2 that I've listed above.  There is a pretty large number of those people around today.  Those people don't really involve themselves with places like Renderosity much anymore.  They have their own proffessional subcommunities.  There are user groups for proffessional Rhino people, 3DS Max, etc..  We've lost those people to their own professional places and we certainly miss them!
Some other people got tired of it, learned whatever they wanted, and sort of fell off the cliff (sorta).
And finally, some of us stuck with the hobby side of it.  The majority of the casual hobbyist though is more often interested in quick results that look amazing without a lot of effort.  Here is a recipe that usually produces great artwork.
i.)  A skilled user.
ii.)  A creative mind.
iii.)  Time and patience.
Most casual people lack in one or more of these three things and consequently, software developers have done their best to provide for those issues but it's not easy.
Before Carrara could import Poser and Daz figures, the artwork that people produced was quite diverse and spanned a lot of topics.  Now, it's mostly just renders of women (which is the vast majority of hobbyist C3DA).
Daz is catering to this obvious market.  It's quite black and white.  We (the Carrara community) should be glad that Daz has decided to keep Carrara alive in the first place.  It's just how the hobbyist world has evolved and what seems to sell.
I of course would like to see a lot of things added to C which aren't related to the human figure.  I don't render people.  I don't care for it.  I have no interest in it.  However, most people want to render women and that's what we have to live with.
When I look at Terragen and Vue, it would seem to me that there is a lot that could be done with C to improve the atmospheric modeling.  I would like to see some simple improvements in the metaball modeler.  I would like to see more controls for weather and clouds.  The texture engine is pretty major powerful but it could use a few more things.
Luxrender is an interesting addition.  I would say that I still haven't really seen things out of it which really show it's value yet but it's there.  I think the REAL power in the luxrender engine involves transparencies and caustics which I haven't really seen a lot of rendering that truly shows it off just yet.  It's there though and it will come.
In the meantime, people in the Carrara community come and go.  Those that have needs which stretch out beyond C leave C for other horizons.
How would we keep those people?  It's a gigantic question.  Here's an interesting example of how people fall away though.
I finally decided a couple of years ago to do a serious outdoor scene.  You can see it in my gallery (titled scout).  In my journey to figure out how to render that scene, I had to do a lot of learning on my own.  If I were a newb to Carrara, it would have been a pretty steep learning curve.  Compared to packages like Vue, I consider my result still lacking in a lot of ways.
However, the outdoor scene people can buy Vue for cheap (to start anyway), they can get Terragen for a nice price, and they can crank out renders of outdoor stuff which is very impressive.  Can C do it too?  Yeah (sorta), it can but you can't figure it by the manual.  You're very likely going to have to get some help to get C to really crank out something.  When you do get help, it doesn't take too long to realize that it takes a lot of little tweaks and dials to get C to really get into a condition to render what you want.
Carrara's true strenghth is that it can do everything.
Carrara true downfall is that it can't do one thing super well.
All of what I've said here is arguable to a large degree but it's just my viewpoint.
What is it that we need out of C9 anyway?  What my gut tells me is that it's going to be some bug fixes with the Genesis stuff in it and that's basically it.  I'm not losing sleep over it.
I can remember when GI was announced as a new rendering feature in C and we were seeing litst release preview renders with it.  It was SO cool to see those big changes coming and we would chew our nails waiting for the release.  Now, I don't know when C9 will come out and I'm not concerned.  It's not very likely to change what we are rendering now.
I think ultimately, what people really want is photorealistic renders and C needs some pretty big new changes to drive to that.  Luxrender is a start but it's not enough.  If we saw C developing towards those sorts of ends while also including all the Daz figure stuff, I think the C community would grow substantially.  Will that happen?  I don't know.  We'll see.

-Kix