Combining the output of two or more antennas will inevitably lead to constructive and destructive cancellations / summing. Indeed the very science and design foundation of many antennas rellies upon using such “phased arrays”.
For carefully spaced, idetical antennas with carefully matched feedlines, the combined performance is both predictable in theory and easily confitmed experimentally. Additive gains at some channels / frequencies occur with inevitable losses at other channels or directions of arrival.
No $20 ‘combiner’ truly exists to magically blend two RF signals so as to achieve gains without commensurate losses. Increasing the physical size of the array / ‘aperture’ by stacking vertically can and will add gain, but compress the elevation beamwidth, gathering received energy in a smaller and flatter torus. Antenna makers at UHF commonly stack bowties to accomplish this.
The real vodoo if any is the behavior of the propogation path. The so–called Beacon Equation and antenna pattern(s) determine the raw signal strength, but terrain, weather, foliage, multipath reflectors, etc. have the truly complex contribution to predicting performance accurately.