I don't think it's a big surprise that all materials can leak after welding.
What's different about aluminium from most other metals is that it has almost the worst possible combination of factors to produce complete fusion to the root of the weld at every section along a seam with no porosity, inclusions or other defects which may result in leakage (not necessarily evident until it's been in service
for some time.)
These difficulties are compounded wherever several welds meet at a corner.
The characteristics which make it difficult include:
an oxide layer whose melting point is so much higher than the parent metal that you'd just about have to vaporise the metal to melt the oxide
a crucial inability of the parent metal to keep a compact weld pool, because the rate of heat conduction is so high, especially in comparison with stainless. The devilish dilemma arises because of a deficiency in 'hot strength': a pool which is even a smidgen too big cannot hold itself in place against the pull of gravity - and melting happens with no colour change, so it's a proper bugger to predict.
One characteristic pushes you in the direction of lots of heat (which would in any case cause unacceptable distortion) and the other pushes you towards flirting with insufficient heat for full fusion to the root.
It's also "hot short", ie very weak during the phase when it's hot but not molten. This makes it extremely prone to cracking as it cools.
It's also a metaphorical 'magnet', in the molten condition, for gases, to the extent that even an inert gas shield is no guarantee against bubbles of porosity in the weld.
These difficulties, while very serious, can be overcome, but virtually none of them apply in the case of stainless.
It has its own difficulties, but these don't make it difficult to weld, provided you understand what parent metals to use, what electrode and filler rod to select, and have sufficient flow of suitable shielding gas.
The actual welding is a delight; the metal flows like water where it's molten, levelling almost like paint
, but holds together really solidly in the immediate vicinity.
A good welder with good eyesight and fine motor
control can be taught to TIG weld stainless to a high standard in a few hours.
The same person would need to practice for months to get results half as reliable in aluminium TIG welding, for leak-free seams.
Modern synergic pulse MIG can be used in thin material, but it's not a rescue
package for a welder who doesn't already have lots of chops.
I'm not saying it can't be done, but you DO have to select a fabricator carefully.
And it's wise to specify a thickness which is on the generous side. By the time you get up to the sorts of thicknesses that hulls are plated from, the difficulties are considerably reduced ... which is just as well, otherwise aluminium boats would not be viable.
And to be extremely vigilant for the possibility of poultice corrosion where it's mounted, as others have pointed out.