I think you might be reading a bit too much into the joke, which is the idea of a scientific paper on giving carbon credits to people conducting actual industrial sabotage, a hilarious concept in itself.

But taking it more seriously, I suppose the argument could be made that delaying large amounts of carbon from being released means reducing X amount of time that carbon in the atmosphere has to contribute to warming and potential feedback cycles. Producing something in a different factory may take time, and while the same amount would potentially be emitted at the new factory, delaying it may not be entirely useless (at least, in my uneducated intuition!).

There are too many variables to know with absolute certainty if a particular sabotage action is overall carbon positive or negative based on how much extra carbon would be emitted to fix the sabotage (depends on the type of sabotage). But if the sabotage results in that production not occurring at all due to making the whole ordeal more costly, it would likely be overall a positive carbon action.

That factory is likely already at capacity and might have to re-tool machines to accommodate a different product, as well there might not be another factory as maybe the sabotage is a crude oil pipeline and sabotaging it effectively halts the operation until repairs are made. Sure some carbon will be released doing those repairs, but that would likely be far less then any ghg's released during normal operation.

Also, burning down the forest before it can be clear cut could actually be a good idea. The British Columbian wildfires of 2022 were likely caused by excessive forest preservation by putting out any and all wildfires for decades upon decades. This made the forests virtual tinderboxes with all that extra undergrowth. Forest fires are a natural process, so causing one to prevent clear cutting is again arguably better then the clear cut, environmentally speaking.