I've run some tests with two types of types of pistol magazines. Both tests lasted several years, one is still running now and is about 3 years along.
Mags that stay fully loaded will lose some spring strength but so far none of them have lost enough to malfunction. I took the mags from the first spring test and all of them fed properly and even locked the slide back every time even after having been fully loaded for over 800 days
Mags that are underloaded by two rounds lose noticeably less spring strength than mags left fully loaded.
Mags that are cycled but not left loaded lose noticeably less spring strength than those left loaded.
Unless the mags get very hot or visibly corrode, I can't think of any reason that a decent quality magazine would degrade in any way from being stored unloaded.
I only tested two magazine types, and only pistol mags, so I wouldn't assume that these results apply across the board. The best answer to this question is to check your equipment periodically to make sure it still functions.
This supports what I've been saying for many yeeaarrrs now and have posted tech data several times to support what I've said.
Springs degrading
only from being cycled is a parroted myth.
If it were true, we'd be seeing recoil springs being replaced at a 10:1 to mag springs... at the very least.
Leaving them fully loaded can and does effect them.
How much of either is dependent on many factors of which probably the most influential in a pragmatic sense is how far is it compressed in to its elastic range. (Think high capacity in compact designs; not full size 1911s, BHP, full size glocks etc)
I say pragmatically because Most gun springs are made from pretty darn similar materials, and temperature isn't too much of a factor because we humans shoot in a fairly narrow temp range as compared to the springs material temp range. Thats not meant to take away from the post above about accelerated life test but to point out that 50C is only about 122F. I've shot at that ambient temp many times when I lived in western AZ. 60C is about 140F. The accelerated life tests I've posted I think started around 75F as the base line and then around 150, 180 and 210F as the data points. And time does play a part in it.
The engineer I worked with at Hughes Space and Communications in El Segundo CA basically said that, general speaking, coil springs only compressed about 25% essentially have an infinite cycle life.
In a practical sense, this is played out in real life with engine valve springs which are compressed no more than 1/4" (that's a pretty agressive cam) on a spring thats around 2" long so we're talking only around 10-15% compressed on a pretty robust spring.
My GMC 5.3L truck has over 500 hrs on the motor (yes, it has an hour meter). If I'm averaging 1500 rpms X 60 minutes in an hr = 90000 X 500 hrs = 45 million revolutions aka cycles on the springs in temp range of... don't known what the temp is under the valve cover is but... maybe 175 F, maybe more.
As a practical experiment, go get a valve spring at the auto parts store if you can even find one because they pactically never fail... or a mag spring, and compress it 95 to 100% and leave it there for a week and look at what happens. It probably won't be functional in its respective application.
Enough rambling and on to the OP questions
If you're in AK I'm going to guess fairly near the ocean and some humidity so spraying them is probably a good idea for longer term unmonitored storage.
The mags shouldn't need to be disassembled but it can't hurt.