On balance, and in all fairness, serious issues with scales are very rare on full tang knives otherwise nobody would be using them, ever. If failure was commonplace folks would take the more reliable solution every time.
I have had one knife returned because of this problem in over 20 years of making them, which was a week or so ago, and it is pictured above. In this case there was no ill will as a result of the material lifting since the knife is a prototype of a design in progress and no money has changed hands. If this was a client project where a final sale had been completed the story could well be very different - I would potentially have a dissatisfied customer on my hands.
My own preference is hidden tang and I make no secret of it, but it is dead easy to introduce problems to hidden tang construction (particularly when peening the tang at the end) which, instead of securing the handle material can do the exact opposite or even overstress the material to the point where it fails.
Bottom line - I advise clients based on their requirements but if they choose a full tang then they won't ever be happy unless they have one, so that's what they get.
In theory both construction methods can produce a good, strong knife. Both can also be thoroughly loused up and both can fail miserably due to any one of a number of factors, most of which are caused by the maker and some of which are caused by the materials themselves or environmental extremes.
The strength factor is academic under normal use and would only ever make itself felt when subjecting the knife to extremes beyond the normal scope of daily work.
I still say that full tang knives have the market share mainly due to economics. Steel is inexpensive to buy but costly to work, so why machine excess away and work on a more critical fit on a hidden tang knife when you can simply flatten the tang and inside edge of two slabs and get the whole lot assembled.
Same argument applies to Scandi grinds - what's the quickest and easiest grind to do ?
Economics governs a lot more than most give it credit for. A fringe benefit is the Scandi grind is good for carving, but the main underlying rationale behind it is very probably down to it being the quickest, least labour intensive method of construction there is.
It is far too easy (and usually inaccurate) to categorically state that X is better than Y.
Seriously - how much can a 3mm or thicker knife flex under normal use to compromise the tang-to-scale joint ?
All else being equal you can good or bad results using either format and full tang knives do not encourage scale/handle failure by default.
I could cite a bunch of very old (over a century) full tang knives I have as examples, and I have a couple that I made about 25 years ago now that are still going strong with no failure and no sign of any.
I keep coming back to the handle material or perhaps the chemical residue in the stabilising process in relation to the wood as a common denominator.