In this paper we address the general applicability of the method pioneered by \citet{Zhang2012} in which the motion of the compact object can be tracked using wind X-ray absorption lines. We present the velocity measurements of the thermal wind lines observed in the X-ray spectrum of a few low-mass X-ray binaries: GX 13+1, H 1743$-$322, GRO J1655$-$40 and GRS 1915+105. We find that the variability in the velocity of the wind lines in about all of the sources is larger than conceivable radial velocity variations of the compact object. GX 13+1 provides a potential exception, although it would require the red giant star to be massive with a mass of $\approx 5-6\ M_{\odot}$. We conclude that the variability of the source luminosity occurring on a time scale of days/months can affect the outflow properties making it difficult to track the orbital motion of the compact object using current observations. Given the intrinsic variability of the outflows we suggest that low-mass X-ray binaries showing stable coronae instead of an outflow (e.g. 4U 1254$-$69, MXB 1659$-$29, 4U 1624$-$49) could be more suitable targets for tracking the orbital motion of the compact object.