A bit of a thought experiment:
The desk top is to be 30" wide, as a guess. So you will have 3 pcs of wood - two each with a single miter across that 30" and one with a miter at each edge.
You fabricate in your shop that has reasonable heat, and maybe little dry in humidity. The wood - Sapele - arrives at 6% MC, and you make excellent, accurate miters and glue and fasten the tops together. Or you leave them loose to assemble on the job - with or without the glue.
The client's house is kept at a fairly high Relative Humidity - say 40% - and within a year or two, the Sapele has grown to a 9% MC - 3% gain.
This means (via Shrinkulator) that the 30" wide tops are now 30-1/4" if flatsawn, or 30-1/8" if quarter sawn. They should all move the same in width. Not a whole lot, but movement all the same.
What happens to the miters? As the wood expands - both halves of the miter equally - that perfect 45 degrees you cut will become a bit less than 45 degrees, and the tops will try to open at the toe or long points. The "U" shape will no longer have parallel sides.
Okay - you anticipated this, so you splined, biscuited, and dowelled the miters, and added a 3/4" Birch ply reinforcement plate on the underside of each miter, with glue and screws. The finish is equal on top and bottom. So movement should be stopped, right?
Well, if the movement is stopped (not always easy to do) - the wood at the miters cannot swell, the miters hold, your concern pays off and the joints are tight.
But the homeowner travels to Europe for a year, and the housekeeper decides it is always too damp in the house and sets the humidity for 15% to better match her desert homeland. After several months, the desk tops will show cracks along the grain on either side of the miter - may be in the wood or on glue joints, but cracks the same. The miters will be relatively tight, but the crack will lead into the field of the tops a few inches or more.
Why? The wood did 'swell' when it absorbed moisture, but it could not go anywhere - could not swell at the miters, so all the fibers all crushed just a little bit all along the joints while they tried to swell.
Then, when it dried out, they retained that crushed fiber profile, and the cracks opened along the grain, radiating out from the miter until they get away from the fixed miter joint and can physically expand and contract.
Solution? I have my thoughts, and I know how I have dealt with this exact same problem, but I'm curious how many readers agree with me, or disagree or can add to my thoughts above.