There is not one problem with the canard layout but a whole rag-bag of problems.
Firstly, it is worth noting that the reason the Wrights built their first Flyers as canards was to make them deliberately unstable. They believed that a plane could not be both stable and controllable; a stable plane would just settle on its course and it would be difficult to change course (This was quite a common belief and was not dispelled for some years). Since they were after controlled powered flight, they went for instability. This they achieved by adding a canard elevator with no extra angle of attack. It was widely recognised that such a flying machine required extreme skill and was too dangerous to take to any speed or height.
When Horatio Barber subsequently introduced his ASL Valkyrie in 1910, he gave his canard that extra angle and stabilised it. It was a safe enough plane by the standards of the day, which is to say no attempt at aerobatics and strictly avoid going near the stall. But it proved a one-off product, with the French style of tail proving much easier to adjust on a prototype and get the production version acceptable.
Many of the design issues arising with a practical canard layout are described in the associated Wikipedia article. If the foreplane carries significant lift, it is likely to stall before the main wing and may disrupt the airflow over the latter. Not only is the main elevator now stalled and useless, but the wing is also losing lift. Such a situation may not be recoverable. On the other hand, if the foreplane does not contribute to lift then the main wing needs to be stable in its own right. A stable wing has other problems, which is why tailless aircraft were also rare. Even if the stall is benign and the foreplane simply drops, pitching the nose down and initiating its own recovery, there is the risk of a phugoid cycle developing in which it bobs up and down with ever-increasing energy until something goes horribly wrong.
Later experimental types such as the FW Ente and MiG Utka were relatively docile, having large enough surfaces with low enough loadings to avoid the worst of the problems. However they were somewhat slow and bumbling. While these experiments might have been developed into practical designs, the necessary research appeared pointless when conventional types were already perfectly adequate.
As engines got more powerful and planes flew higher, faster and further, the potential benefits of the canard led to a re-examination. The combination of good manoeuvrability with a tame stall and high efficiency proved elusive. Types with complex high-lift control canards, such as the Curtiss XP-55 Ascender, sought to avoid these issues but it turned out the canard would still stall under some flight conditions, and when it did so it was lethal.
The canard received new interest in the supersonic era, due to its potential for a smaller wing, higher manoeuvrability and a better CG range than a pure tailless swept or delta wing. But, even with a thin delta wing on something like the North American XB-70 Valkyrie, the core stability and control issues remained. (Concorde would not need much manoeuvrability and got over the CG issue by pumping fuel to and fro, so it would abandon the canard).
Torsten Örnberg at Saab realised that these problems could be overcome by close-coupling the canard to a delta or swept wing, using the canard to positively affect airflow over the main wing, and in particular to create and stabilise lifting vortices at low speeds. From his US patent, filed in 1963:
"This invention relates to aircraft of the delta wing canard configuration and refers more particularly to improvements in aircraft of the type having a thin, sharply swept-back main wing and a secondary wing located ahead of the main wing, which improvements overcome certain heretofore existing problems relating to the stability and controllability of such aircraft."
Part of the reason this worked is because the swept and delta wings are relatively easy to make stable in their own right, and a great deal of work in that area had been carried out during and after World War II. The patent also discusses various problems with other canard arrangements, so I am not going to repeat all that in detail.
What I do suggest is that Örnberg's innovation was a remarkable creative insight into a tangle of issues that had defeated every aeroplane designer since Horatio Barber. Usually, such close-coupling of obstructions in front of the wing causes at least as many problems as it cures, and it appears far from obvious that it would work this time round. I'd go so far as to call it a stroke of genius, and its success after half a century of failure a stunning achievement. Its significance also went well beyond its immediate application to the Viggen, the first canard to enter production since 1910. Suddenly the inhibitions of half a century fell away, other designers began to understand how to think about canards and they started appearing on new combat types and being retrofitted to several derivatives of the Dassault Mirage III. The Viggen inspired Burt Rutan to start a similar revolution in civil aviation. Not bad for a relatively small country which seldom sold its warplanes abroad.