I have retuned many in the field and in the driveway including PD, ERI ring-stub (Collins), MCI twin-V, Jampro penetrator, Bext and other LPFM ring-stub types as well as yagi arrays for translators.
If it has full wave spacing and hard-line interbay lines, its pretty easy.
If it has a power divider and flexible lines its not hard if you can get to the bays easily for tuning.
If its not full wave spaced and has hard-line interbay lines, I would not try it. Also directionals are not on the menu.
ERI rototiller antennas are supposedly good for about 4 mhz (108-104, 104-100, 100-96, 96-92, 92-88) without changing anything except the location and number of the "donuts" in the matching section.
I can't help you with that, since the tower crew does that after install based on Network Analyzer data and Smith chart. You need a "real" engineer for that.
If it has hard-line interbay lines and a stub-matching section __I__I__I__ at the bottom you can probably just adjust the bottom and middle stub for best match (for a 1.5 mhz move) with pretty good results. Use an exciter, and a Bird wattmeter to get close and then tweak the bottom stub for best match at full power.
If it uses a power divider and flexible lines to the bays, you will have to retune each bay (usually for 50 ohms +/- j0) to get the impedance right at the input to the power divider. If you can do that, then the length of the lines is immaterial (as long as they are all the same length!).
If you can take it off the tower and rebuild it on the ground you can get the best results. I mount it on sawhorses about 2 feet off the ground with the bays up. Hose clamps, vise grips (careful!) and woodworking clamps are very helpful in getting things to stay in place.
Hard line interbay lines need to be adjusted for length to prevent unwanted "beam tilt" (if you are moving very far).
A move UP in frequency requires shorter, and a move down requires stretching them (replace with longer). Air lines have a velocity factor close to 1, so (11,820 / Freq in mhz) gets you a full wavelength spacing (in inches) between where the bays attach. For a thorough job, the bays need to be adjusted also. Use the above formula, then divide by 2 to get a half-wavelength for the bay elements.
Measure everything before changing, compare to the existing frequency, then adapt/check that ratio to the new frequency. The calculator is your friend.
If they have a stub tuner, adjust your loops, lenghts, etc. in unison for the best SWR without the tuner attached. Then the tuner can be tweaked to get a good 50 ohm match for the coax.
I use the MFJ impedance bridge for this work when I am away from other transmitters, but it is useless if there is any other transmitter operating nearby. Generally speaking, the length of the radiating elements will determine the frequency where reactance is closest to zero (resonant frequency), and the length and location of the loops or feed connections will determine the resistance.
You can also use the MFJ to get a good idea of bandwidth by checking how far you can go before the SWR rises to say 1.5:1.
An agile exciter and bird wattmeter can also be used for this.
Good luck.
