Lacking additional information, answers likely will be speculation-based only. To mention a few factors to consider:
Stating GC-MS analysis only states, that there is technique to separate compounds by different rate of retention over a stationary phase (gas chromatography, GC) combined (a.k.a. hyphenated) with an other technique to analyze fractions sorting at different times by mass spectroscopy (MS). Mass spectroscopy typically shatters the molecules into charged fragments which, by their mass/charge ratio, behave differently in electromagnetic fields. However, depending on the setup, the separation of the fractions by GC may be better, or worse. On the side of mass spectroscopy, there are multiple methods to ionize and fragment molecules and to record these. It is a problem that the parameters of running a mass spectroscopy may alter the shape of mass spectra quite a lot; signals may vary in their relative intensity, or sometimes be below the detection limit altogether. It need not be the engineering advantage among corporations, of Agilent's technology over the by Bruker (if there is?).
In the line of smaller benchtop mass spectrometers typically mounted behind a GC, larger sector field mass spectrometers, and Fourier mass spectrometers, there is a significant increase of resolution to record tiny differences in the masses of the fragments the experiment of mass spectroscopy generates. This may help to attribute the signals to the correct analyte, especially if there already is a suspicion to search for a selection of compounds only.
Routine GC-MS used for forensics and IOC doping screenings, for example compare the mass spectra recorded for a sample against previously recorded trustful references. (Different to IOC doping screening, it is not evident if the Russian government shares / shared such standards here, though.) Depending on the compound to identify, though, the analysis does not need to stop at the level of a GC-MS. Both other types of chromatography may have been used now (e.g., HPLC) to separate the compounds; or multiple mass spectrometers coupled into a tandem (MS/MS) to observe not only the initial fragments obtained by mass spectroscopy, but equally their further fragmented «daughter radicals», too which equally rises the chance to identify your compound in question.
Maybe the Germans did not trace the poison itself, but metabolites judged as sufficient evidence for earlier reported, related compounds. Especially with small amounts of sample material, the success of an analysis equally depends on every step of sample preparation; here likely (incomplete listing): what was the delay between Nawalny's intake of compound X and obtaining a sample by him (metabolism), the nature of the sample (tissue, blood, hair, etc.), delay and conditions of storage of the specimen to get it to the lab (unwanted sample decomposition), concentration and digestion protocols in the lab to eventually submit the sample to the mass spectrometer; the people's experience and availability reference data / techniques.
