In this situation, "compatibility problems" seems too vague a topic with which to work. The hypothesis would require a specific compatibility problem and study a different company in order to be valid. This will help to keep the dissertation manageable (Rockler-Gladen, 2007). The fourth topic "The effect of bonus schemes on staff turnover rates in the countries in which my company operates" would, more or less, be valid. The country or countries should be specified. Otherwise, with a proper methodology developed, a hypothesis could be developed and tested, making this a valid dissertation topic. The topic is easily quantifiable and can be researched. The issue is contemporary and relevant both to academia and to industry. There will also be literature that can help to support the dissertation. Therefore, it meets the criteria of an acceptable dissertation topic.

The fifth topic, "An appraisal of Microsoft Office compared with its competitors" would not be a valid dissertation topic....

Such a topic is open to bias in terms of the selection of metrics. In addition, this topic is more suited to a consumer review magazine than an academic journal. The subject is simply not of relevance to the academic knowledge base. Every consumer has his or her own evaluation criteria. Moreover, there is no causal or corollary relationship to be tested. Ideally, a dissertation would test the impact of one factor on another factor. In this situation, that does not occur. It is merely a subjective evaluation, rather than an objective testing of a hypothesis that can be supported through the gathering of quantitative evidence.
Works Cited:

Moore, B. (2009). Choosing a dissertation topic. EZine. Retrieved May 7, 2010 from http://ezinearticles.com/?Choosing-a-Dissertation-Topic&id=3055325

Rockler-Gladen, N. (2007). Dissertation and thesis topics. Suite101. Retrieved May 7, 2010 from http://graduate-schools.suite101.com/article.cfm/dissertation_and_thesis_topics