Gede: The bottom line is that they believe that the money they would make with Linux (and Mac) sales will not make it worth the investment on that title.
shmerl: Nothing stops them from crowdfunding such effort, if money is the issue. I assume it's not about the money really.
This really ignores several core issues
Profitability and opportunity cost
Its all about money. Games crowdfund because game development costs are front loaded. You have high up front costs. You are basically time-shifting revenue from the future to today. But games still have to MAKE MONEY. They're not a charity and CDPR as a publicly traded company is interested in profit
So crowdfunding doesn't solve the issue of profitability. You still are going to put X money in, and are getting Y money out. crowdfunding just defers a percentage of Y to the present. But the bucket of sales is still Y overall, and profitability is viewed from that perspective. If the linux port is unprofitable, then its not worth doing.
Then there is Opportunity Cost. A dev studio has finite resources. So resource allocation is zero-sum. Resources put into A do not get put into B. So not only is there an actual dollar cost of X for putting a bunch of engineers on this Linux project for 1 year. But there's the opportunity cost of "if I put X engineers on Witcher 3 Linux, that's X engineers that are NOT working on Cyperbunk 2077". That opportunity cost could be "we cannot implement feature A without those engineers" or "we must delay teh game because we don't have enough engineers".
So now you say "well just outsource the porting" ok so now you've increased the cost by A LOT so your profitably goes down. Also outsourcing doesn't mean 'you don't need to manage the project'. In fact to do so requires a lot of dedicated resources, time, energy from your core team. porting is not a 'fire and forget' thing. Thus again you are sucking away resources from other projects to the port. Is that worth it.
In this situation your profitability question is not just "Are we making money X" but also "is making X money even worth it compared to the lost resources to other more profitable projects". As a core example I'd like to point out Stardock's Elemental Game vs Impulse Store dilemma. Elemental was a disaster mostly because of Impulse. Impulse store was highly profitable and thus the company re-allocated limited resources to the thing that was making money. The question of "where do I put engineer X" is easy when you see Impulse is making boat lods of money but not Elemental so the opportunity costs of engeineer x is better spent on impulse than Elemental.
Thus the question is not just "we wil make $100k on Witcher 3 Linux" is also "was that worth it because we probably could have made $1 million on Cyperbunk 2077 with those engineers". Thus even making money can be considered a LOSS with regards to resource utilization.