No way. To justify this you would need to be using a huge amount of nitrogen. I've looked at these things for laboratories . Still hard to justify . I suppose if you had a junkyard full,of free parts and liked these kind of projects, sure. You would need access to cheap power and a darn good compressor.
Keeping it dry, that's a whole big problem on it's own.
Did you look at the link and also consider my comment about scaling down the system?
Here's another link with more details about the system he built.
http://homemadeliquidnitrogen.com/PSA/index.html
He used a 6hp compressor at a flow rate of 6 to 9 scfm. In his device he pressurized the adsorption beds to 110-120 psi. A lower-flow system with enough flow to continuously purge a box would need only a tiny fraction of that power and flow capacity. I assume that 90 psi would probably be enough to operate the adsorption beds. This is easily achievable with a small 100psi compressor for under $60, such as a 1/3 hp unit from Harbor freight that is rated at 0.6 cfm at 90psi, which already is far more flow than would be needed to purge an enclosed cabinet to prevent oxygen infiltration. For example, if one only used 0.06 cfm then the system could run at about 10% duty cycle. Assuming an efficiency factor of 50% for sake of discussion, that would be enough to supply an average of 0.8 Liter/minute. One could probably get away with an order of magnitude less flow than that, which would bring the duty cycle of even that small compressor down to 1%.
Since this is a non-critical application, I think one could make most of the device from mild steel parts rather than stainless, which would drastically reduce the cost.
The person who built this device also included measures to remove moisture and particulates. I think a first stage of moisture removal could probably apply at the compressor air storage tank by adding a small bleed valve at the bottom of the tank. This could bleed off condensation of much of the atmospheric water that would occur when compressing the gas.
By the way, I have also looked into nitrogen generators in the past, in this case to supply nitrogen for tandem mass spectrometers. I no longer work at that facility, but as far as I know they never adopted nitrogen generators and are still using liquid nitrogen boil-off from a huge liquid nitrogen tank to supply nitrogen for dozens of tandem mass spectrometers.