We recommend that you use "PRM" instead of "Targeted" now.
With "Targeted", for any particular spectrum, Skyline finds the one precursor in the document whose m/z most closely matches the precursor of the spectrum. This would sometimes lead to unexpected behavior where everything works fine for a small Skyline document, but when you add more peptides to the document, some of these new peptides end up stealing spectra away from other peptides, and those peptides end up with empty chromatograms when you reimport the data.
With the new "PRM" acquisition method, a spectrum will contribute to the chromatograms of all precursors whose m/z is within the "Method Match Tolerance M/Z" (the setting at "Settings > Transition Settings > Instrument").
You will probably find that the "PRM" acquisition method works perfectly for you. If it does not, and some precursors are getting chromatogram points that they should not have, then you might need to lower the Method Match Tolerance m/z.
(If you really need to have the original behavior of the Targeted acquisition method, we could release a new version of Skyline which un-hides that option, but so far, the new PRM acquisition method has worked for everyone).