There are several potential causes for reed pitch drift, or just one off changes, some of which have been mentioned above. The pitch is set by the combination of the 'spring' in the belly of the reed in combination with the mass of the moving reed tongue section. It is also influenced by the damping and reinforcement from sound wave reflections resulting from the dimensions of the acoustic chamber. Finally the air flow through the system, from the reed tongue gap, out past the open valve and through the open pad hole. The last variable is the mechanical couple between the reed frame and the reed pan, how tight/ loose it is. I also ought to mention the stresses in the spring area of the reed can affect pitch if they are re-distributed through the reed section
We file the belly of the reed to flatten the pitch, and we file the tip to sharpen the pitch, so clearly muck on the tip will flatten the reed, or corrosion pitting in the reed belly near the reed root can weaken (flatten) the reed too.
we know that a reed sound about 5 cents sharper out of the concertina than when sounded in the assembled concertina, this is the effect from the acoustic chamber, I have been told by very respected individual that if the pad lift height changes the reflective characteristic from under the pad and can influence pitch. I can't say that I have ever come across this myself, but I report it for completion's sake.
Air flow, or resistance to air flow, however, is very obvious. As valves stiffen with age then they can have the effect of dampening reeds, equally if the tongue set or gap is wrong the reed pitch as well as response can be affected.
When you file a reed you take off a surface layer, that surface layer held stresses which maintained the shape of the reed in the reed frame. File the reed and it curls away from the surface you just changed. We can talk about the neutral axis shift but please just accept that it happens. The curl closes up the reed gap so you need to carefully bend the tongue to re-gap the reed. This action takes the steel through it's elastic limit and re-sets the stress distribution in the reed. by so doing you have now slightly flattened the red again.
However the most common cause over time, is the phenomena of work hardening and age hardening. With continual vibration and springing the crystal structure boundaries of the reed tongue start to be more fragmented causing the metal to harden a little. this makes the reed 'spring' stronger and more difficult to move under the influence air flow. A situation which can flatten the reed pitch. Work Hardening can also initiate micro cracks and hence metal fatigue. Brass is more susceptible to age hardening but steels can change in time too.
So yes, reeds can change pitch with time, a long time. This drift is usually through muck or corrosion, It can also be metallurgical but in steel reeds this is usually minimal. Most of the pitch influences are set by the instrument build design, acoustic chambers, materials used, pad hole diameters etc. so they are constants anyway. Then there are the things which we do to reeds that can deliberately or inadvertently change pitch, but that is not pitch drift. Although the playing in after a re-tune can cause pitches to drift a little as they settle down
In my work flow I always clean the reeds first, then replace the valves. Only then do I measure tuning. New valves and clean reeds can reduce the amount of filing needed. I always resist tuning an instrument with old or obviously not too new valves. However, that is a client decision in the end. Finally I usually play the instrument a bit before admitting that it is complete, especially really nice ones, and really, really very nice instruments take longer to play in. A phenomenon yet to be understood.