For many years especially with Agilent 230C was the maximum temperature limit on the source because the equipment would not go any higher without damage to the source itself. This was with the introduction of the 5973. With the 5970, 5971 and 5972, the source was not directly heated, but instead was heated by the interface temperature indirectly so the source was actually only about 170C.
As you increase the temperature of the source you will also increase the amount of background noise present and in most cases you will decrease sensitivity. If you normally run very dirty samples with high concentrations of high boiling point contaminates or target analytes you can increase source temperature to help keep it clean longer, but you have to balance that with the increased noise.
For the inlet, you need a temperature that will quickly vaporize all of your target analytes but not high enough to cause breakdown of those compounds. If the temperature is too low, you will have lower sensitivity due to the compounds not vaporizing and entering the column or they will have tailing peaks because the analytes vaporize over a long period of time. You can adjust inlet temperature higher or lower to optimize performance to each different analysis you do.
Maximum program temperature is also dependent on your target analytes and of course the limits of the column phase being used. Some columns used for volatiles analysis by GCMS only have a maximum temperature of 230-250C so your program can not exceed that limit, some columns like the DB5 phases can have a limit of up to 330C. Depending on the elution temperature of your analytes you may only have to go to 200C, or you may have trouble eluting even when going to 330C, you just have to know your column and your analytes to optimize performance.
Overall the inlet, program temperature, and source temperature are not directly related to each other, but more so to the analysis you are performing on the instrument. On my volatiles instrument I can operate with an inlet temperature of 170C with column program from 35C to 230C and have the source at 210C, but on the instrument that analyzed semi-volatile compounds the injector is 250C, column program is 50C to 330C and the source is at 300C. The temperatures you list are hopefully what has been found to be optimum for that analysis, but you can experiment by changing each to see what effect it has on results, sensitivity and background noise. Though you probably would not want to make changes in the method until it is all approved.
