The answer depends on the atoms or molecules you're analyzing, their sample matrix, and their expected concentration level.  In short you want to maintain your current capability and even enhance it!

Personally I'm partial to ICP-OES/MS.

We analyze a wide variety of samples including food products, household chemistry , sorbents and so on. In our case it is hard to decide on basis of elements and/or sample matrix which method is the best.  That's why I'm looking for users feedback. I realize that ICP-MS is the best possible solution, but the cost of purchase is to high. The costs of operation of ICP are also quite high compared to F-AAS and MP-AES. As for the capability if we need to go down with the detection limits we can still use graphite furnace. However in most cases flame atomization is enough, that's why we consider those three equipment's.

In that case I would continue to use 'flame AAS' but with the ability to mount a 'graphite furnace (for Lead or Mercury analysis) and a 'cold vapor generator (for Antimony, Mercury, Selenium).  I think Agilent or Perkin Elmer have such a system.

It is the extraction procedure that is more relevant to the analysis and the number of samples you receive each year.  In any case maintain or increase your flexibility!

All of these equipment's have increased flexibility compared to "conventional" F-AAS. Unfortunately,  non of them is the one and only best solution, that's why I'm so interested in users feedback.

Exchanging burner with graphite furnace is quite old solution, nowadays most producers offer tandem or stand-alone solutions. As for hydride/cold vapor generation all of these solutions have this option, morover MP-AES has interesting solution caled MSIS where you can analyze "normal" and vapor forming elements at once.

In my opinion, I would choose the MP AES Agilent . The instrument is robust and has a low operating cost. The HR CS FAAS, allows the determination of diatomic molecules as NOx, SOx and POx. If these analyzes are in your routine, HR CS FAS would be good. However use may be somewhat limited.

Last year I bought a 700 AJ ContrAA. I had studied the purchase of other equipment but finally I decided on this one. When it does not work the cost is zero, of gases and of electric current. With the hydride device and the graphite chamber you can determine almost all elements even at very low levels. Even F as CaF, and S as CS, for example. It also has a direct solids analyzer (without previous digestion) that is unique in the market and has many possibilities. It's certainly a equipment for analysts looking forward to doing new things, not just routine work. Personally it was my choice.

Do you have any previous experience with zeeman corrected graphite furnace? Are the detection limits similar for regular procedures?

I'm aware of the solids analyzer but it's quite expensive. How about sensitivity compared to digested sample?

My previous equipment was a Perkin Elmer 5100 Zeeman. The limits of detection are similar in some cases and in others they improve. But there are other parameters that improve a lot because there are really important technical changes. I advise you to consult with a commercial service Analytik Jena. And look at the web page. It's true that the solids analyzer is expensive, but there is a fairly inexpensive manual option.