One of the clear benefits for undertaking latent growth modeling or curve of factors modeling is the ability to operationalize the actual change in our variables of interest and to examine what may directly impact those changes and/or how those changes impact other outcome variables.  This is clearly superior to cross sectional designs.  However, you can only study change in your variable if there is actually something that may trigger that change.  Specifically, if you simply go out and collect data on multiple occasions from the same set of individuals, there is no guarantee that change will be detected in your focal variable.  Why you may ask?  Well, it may be the case that there was no triggering event to promote change in that variable.  

Why am I mentioning this?  It's simply that if you are truly interested in studying the dynamic/change aspects of your focal variable, you need to carefully select your sample and the context against which that sample resides to increase the probability that change will occur.  In short, you need to be purposeful (see the Ployhart and Vandenberg Journal of Management article for more explanation on this) and creative.  This requires you to think of scenarios where you might actually observe change.  For example, if you collect data from newcomers to an organization, there is a higher probability that you will detect change as those individuals become socialized/acclimated to their new work place.  Another example is an organization that is struggling to remain competitively viable.  You would expect individuals there to be changing their views of the workplace in that context.  Intervention studies such as the ones I've been involved in there is certainly an expectation of change in the variables aligned with the intervention.  

My major point is that simply collecting data from the same sample on multiple occasions is not going to guarantee that change in your variables of interest may occur.  You need to be strategic in your research design considerations to increase the probability that change will occur.  

Please let me know if this makes sense or not.