Texas actually has more registered Dems than Repubs, and Repubs need independents if they want to win statewide.
There is no party registration in Texas, so I don't know where you're getting this from. 2004 exit polls (for what they're worth, I know) had Party ID @ R 43% (+1% from 2000), D 32% (-8% from 2000) and I 24% (+7% from 2000)
Chris Bell is, repeat after me, one of the worst candidates, period, that anyone could nominate for anything. I lived in Houston during the period where he ran for city council (successfully), mayor (unsuccessfully), House (successfully and then unsuccessfully). He only has appeal to white liberals in inner-city Houston (they do exist). He has zero appeal to Democratic minorities, especially blacks, and will only get their votes in this election because their is no minority candidate on the ballot and he has a D by his name.
One other thing that I would add. Texas is already much more like a 55-45 GOP state than a 60-40 GOP state and has been. That being said, it is fairly stable for now (and has been since the mid-1990s). Even in the Dole disaster of 1996, he still won the state by 5%, while Clinton was winning nationwide by 9%, a margin of 14% total. The Bush margin has been more like around 20% from margin of national victory, but I would still place the Bush effect at roughly 3%-5% (I think there has been a tiny shift towards the GOP in presidential contests only)
I have said countless times on this forum that the race to measure partisanness in Texas is the Cornyn-Kirk race of 2002, an open Senate seat, with two clear representatives of the bases of both parties (without a Bush factor to take into account). That race was 55%-43%.
Whilst I do agree with you in terms of the "too much power for one party" theory, I simply don't see it coming into effect, yet. As jimrtex has pointed out countless times, it is very hard to campaign against incumbents in state Senate seats because of the ground to cover in campaigning and the money to make a viable campaign. (only 31 Senate seats as opposed to 32 CDs). State House is much easier and changes could happen quicker there, but still a lot of seats are uncontested right now. Anyway, we'll see this fall.
Anyway, my main point is this in what the future could be in Texas, looking at what I've noticed in Texas in the past 20 years or so: I have noticed that we continue to see GOP degradation in the inner suburbs of Texas big cities, much like in other major urban areas, though less so than outside of the South.
Dallas is pretty easy to figure out. Houston is actually mainly Fort Bend County, but a lot of this has to do with the no-zoning Houston policy. Austin is predictable, but the shift is less rational, as it moved the other way in the 1990s, San Antonio less than the other cities above.
Rural white areas moved from voting Democrat to Republican about 10 years ago (east Texas) or much longer ago (the rest). Black vote has not changed, though vote numbers have decreased.
What has kept vote margins stable over the past 10 years is the following: Growth of exurbs (particularly into Mont. Co and Brazoria in Houston), Fort Worth (for Dallas), Williamson Co. (for Austin, though less so, once again). Degradation of Democratic Hispanic voters in rural South Texas and urban areas in Houston and Dallas. The Bush effect skews this way too much, but it is still occurring, just look at the election results. Note (and this is important): The move is towards ID as Independent, rather than Democratic. More important: This vote effect is much stronger in Presidential contests than in local ones. Most of the Hispanics who ID as Indys vote locally Democratic; I lived in these neighborhoods in my many years in Texas and saw this quite clearly develop, even though it's very slow.
What this should mean for the next 10-20 years is that Republicans will fair much better in Presidential contests than they will in statewide contests and vice versa, unless the statewide Republican does exceptionally well with Hispanics or the national Democrat does exceptionally well with Hispanics.
The future success of the Republican and Democratic party in Texas depends on Hispanics more than anything else (which is sort of obvious). But it also depends on the way Hispanic voters identify themselves. The Democrats will have greater future success if they keep Hispanics indentifying and voting Democrat on all levels of government; the Republicans will have greater future success if they move Hispanics from indentifying and voting Democrat at all levels to gauging more as Indys on both the national level and then the state level.
So, that's the long and short of it, in the analysis of the past, with regards to what I see for the future, taking into account the Bush factor and looking over a long enough period of time to give some other factors time to flesh out.