User Experience Design is more than quantitative site analysis - by its nature as part of the design process it involves making subjective decisions on aspects where it is difficult or impossible to get data, and must take into account a balance of factors which make up the experience gestalt.
UX design takes a holistic, multidisciplinary approach to the design of user interfaces for digital products [...] ensuring coherence and consistency across [many] design dimensions. (Source)
This leads to the conclusion that yes - UX design needs to take into account "non-tangible" factors.
To answer your sub-questions (note this is opinion, YMMV):
1) Rollout strategy
All parts of the design team need to be concerned with rollout strategy. For UX this may involve making compromises in the design stages in order to allow the rollout to proceed effectively and efficiently - but more importantly, the rollout itself can have UX implications.
Initial rollouts tend to be to a new user base. In this case, it is the responsibility of the UX designer to identify likely users and ensure that the site or application delivers the appropriate experience. This is a bit of a no-brainer. Further rollouts iterate upon this, and require the UX designer to take new user groups into account (and if the rollout is part of a larger strategy, to plan for this stage in advance).
Transitioning between two different designs usually has a negative impact on the experience - as a rule, people hate and fear change (even if it's for the better). Managing the rollout of an update comes firmly within the bounds of UX, and it is the responsibility of a UX designer to oversee (and mitigate if necessary) any issues caused by changes to design.
2) SME test driving
The answer to this question depends on a few things. Firstly, are you targeting SMEs, and if so, why?
If you're targeting SMEs as one of your key user groups, then yes of course you should test with them (as you should with all other core user groups). This is particularly true if your site's uptake strategy relies on SMEs driving further growth through authority/exclusivity and network effects.
It seems from your question that you're looking at a rollout strategy which includes core users first and then a further rollout to other users at a later date. In this case, since your primary users are solely those in the SME group, then you will definitely need to test with them (if only because they represent the totality of your initial user base).
Remember though that SMEs may not be representative of your final target user base. If you are planning on rolling out the site further, make sure that you can test with your eventual users as well, and if necessary come to a compromise on design decisions.
This question is a bit too vague to answer without knowing your circumstances. It may be something that you should bear in mind, but without knowing your strategy and methods it's impossible to say whether it would be a factor.
4) Staged rollouts and betas
A beta is a very specific thing, and different to the first stage of a rollout to end users. It is a sign to your users that you may not be feature-complete and may not be bug-free. Given this, signalling beta status to your users is a method of managing expectations.
The question is: do you want to subject your core user base (if you are relying on your SMEs to drive engagement and uptake later) to a beta? It's certainly possible to use them as your guinea pigs, but it's important to remember that this stage may drive some of them away if there are serious problems.
In summary - yes, it's probable that staging a rollout will improver the experience for users later in the rollout, as bugs and other issues will naturally be caught as your initial users encounter them. Whether this is good for your initial users is a different matter, however. Further, you need to take care that the needs of your initial users do not have undue weight on your design decisions if you eventually decide to cater to a wider user base with different needs.