On HF, the only "very reliable contact path" is one that exclusively employs surface wave, and thus has no dependence on reflection from a tropospheric or ionospheric layer.
All other HF paths involve one or more reflections by layers of particles ionized by atmospheric conditions or solar radiation. The degree to which a layer reflects an incoming HF signal back to earth (instead of absorbing it or allowing it to pass through and interact with higher layers) depends on the signal's frequency, the layer's composition, the recent history of solar radiation, and the recent history of the earth's geomagnetic field. Time of day, month of year, and year of the 11-year solar cycle all impact the composition of the ionosphere's multiple layers. At any time, a solar storm can unleash a torrent of particles that perturb or enhance propagation on some paths; given earth's distance from the sun, we get advanced warning of such events from sites like
this one.
The best source of information about HF propagation I've encountered is
The New Shortwave Propagation Handbook, available from CQ Magazine.