Strangely enough, Zambia has considered such a plan recently; to increase water supply to Lake Kariba, not for navigational purposes.

Although an Atlantic-Indian ocean shortcut through Central/Southern Africa looks tempting on a map, the logistical difficulties would be frankly immense.

First, the good news. The River Congo is navigable for long stretches of its length. This is the only thing in the plan's favour.

First problem. The Congo is navigable in long stretches, but the Livingstone Falls is an enormous series of rapids-cum-waterfalls which would be eye-wateringly expensive to bypass by canal. Sure, you could build a railway to bypass it (and the Matadi-Kinshasha railway does just that), but then you haven't got a navigable water route any more.

Second problem. The area where the Zambezi and Congo watersheds meet has a decent annual rainfall, but there is a marked wet and dry season. Massive reservoirs could be built, I guess, but again it would be a massive engineering challenge.

Third problem. The Zambezi isn't really navigable either! The obvious hindrance is the Victoria Falls, but this is but one of the many sets of waterfalls/rapids along the Zambezi's length. The situation is better than it was in colonial times, due to the building of several large dams, but even so, enormous lock systems would be required to make the Zambezi navigable from source to sea.

All this makes such a connection enormously expensive, and utterly useless without incredible feats of engineering. It is likely that any colonial administrator of the time would have quickly realised this, and opted for railways instead. Which indeed was done fairly extensively in the region.