The theory is the same regardless of frequency, but the practicality usually isn't, so people chuck theory out the window at lower frequencies -- it's just impossible to follow what theory dictates.
At two meters, theory can be reasonably followed.
Any number of antennas will be fed in phase if the transmission line length from the signal source to each antenna is identical, or if not identical, then precisely a wavelength integer (calculated in the transmission line, using its velocity factor) different.
Matching multiple antennas to a single transmission line is not difficult and is extremely well documented in all the antenna handbooks; although the most appropriate books for 2 meter discussions (where there are more matching options available than on, say, 80 meters) would be the ARRL VHF-UHF Manual or the RSGB VHF Handbook. Both are excellent, I own and update both every few years, and the information, due to copyrights, is *not* on line.
Stacking for maximum gain in theory (at any frequency) is achieved when the near-field forward lobes just meet, and don't overlap, in space. The distance between antennas, to achieve that, is based on the antennas used. For conventional "beam" antennas like yagis, quads, or quagis, it occurs when the antennas are spaced about one boomlength apart for booms up to nearly 2 wavelengths, and then somewhat closer than one boomlength apart for longer boom antennas. The actual optimum spacing (stacking separation) distance may be calculated using most of the popular antenna modeling software currently available. I use EZNEC, mostly. If you don't have a copy of the program, a demo program is available for download for free; however, it's limited. The "purchase" version is more complete.
Using a modeling program, it becomes very evident that optimum spacing for maximum forward gain is not the same as optimum spacing for minimum sidelobes, or for maximum front-to-back ratio. For this reason, most amateurs choose to use a separation that is a compromise between these parameters. The algorithms to model these are so complex that until software became available to do so in an automated manner, hardly anyone bothered to go through the calculations and simply used the "antenna manufacturer's recommendations," which in the past were partly guesses and partly based on antenna range profiling.
If you'd like to do your own range profiling, for the HF bands it's very difficult; but for 2 meters, it's pretty easy. There's a great writeup on antenna optimizing using amateur methods and equipment on the N6NB website,
http://www.n6nb.com.
Dr. Cebik's website,
http://www.cebik.com also contains excellent information on antenna optimization and stacking.
WB2WIK/6