It depends a lot on exactly which molecules you care about. Some proteins, especially phosphoyrlated ones, and some antibodies are known to be highly sensitive to iron contamination. Phosphopeptides and phosphosugars can vanish entirely due to metal contamination. (This is an underappreciated problem with electrospray interfaces.) There are other proteins that simply ignore metal contamination.
Years ago, when I worked for Pickering Laboratories, iron contamination was killing their glyphosate columns. I had to develop a special cleaning solution. Switching to non-metallic frits prevented most of the problem.
Passivation and deoxygenation are good practices to minimize corrosion and contamination, but may not always be sufficient. One of our customers validated an important protein assay using stainless steel HPLC equipment. After years of use, his equipment is shedding iron at a furious rate and poisoning the columns.
Titanium is less susceptible to acid corrosion than stainless steel, but there are several alloys of titanium used for HPLC construction, and they are not equally inert. Phosphorylated compounds do stick to titanium dioxide under acidic conditions (there is even a TiO2 column sold for that purpose).
PEEK systems have the lowest metal leaching. Dionex would be happy if you bought some.
In some applications, for instance dsDNA, it is usual practice to include EDTA in the mobile phase to complex divalent ions that interfere. I have demonstrated that pyrophosphate can mask or remove iron contamination in phosphopeptide analysis. These additives, unfortunately, are bad news for LC/MS.
