database size vs. performance questions

I am attempting to estimate database performance based on a prototype we have developed. I have not attempted to fully populate the database for the timing measurements made (I don't have time to generate that quantity of data :-), instead I was hoping someone on the net could help by answering a couple of questions:

1. How does the size of a table affect performance? (Obviously, the larger the table, the longer it takes to query.) Is there a factor which could be used to translate the response time for a table with 1000 rows versus a table with 100,000 or 1,000,000? Is there a similar effect on insert or update performance? (I did a little test on one table and saw little or no difference between inserts to a table of 1000 rows versus 100,000 rows. Does this jibe with your experiences?)
2. Do changes in the overall database size affect the performance of queries (and updates) of a table whose size is unchanged? For example, is the performance different when querying a 1000 row table that is part of a 20MB database versus a 100MB database (assuming all other factors are equal and both databases are populated near capacity)?

I would appreciate any replies and will summarize if there is interest.

                  Thanks in advance for your help!
                         jim