There are too many reasons for performance degradation, had to list all in a single post.
Let me mention that one of the chapters in the Developer II training manual is devoted completely to performance tuning. I will name a few major causes of bad performance:
1. Use of IF functions in aggregated formulas
2. Use of Date Islands where tables are not linked in the data structure. Any structures with tables not being properly linked, can cause a Cartesian multiplication.
3. Complex data structures where facts are scattered across multiple fact tables, and there are too many links that need to be resolved in the run time.
4. Overly detailed chart, especially Pivot Tables, often in combination with IF conditions.
5. Using overly complex expressions on top of an overly large database. On large databases, expressions should be kept simple...
I'm kind of looking for more practical thiings...like due to joins the script takes lot of time to reload or say due to an exp a particular chart takes more time to show up.
For script related issues I would look at the script log to see what operations take a long time.
When it comes to sheet objects, take a look under Settings > Document Properties > General > Memory Statistics, which should give you an indication of the calculation complexity of the different objects.
