The chief issue with tribalism is that it results in inaccurate/wrong ideas being pushed. This happens due to: people collectively bargaining (which is an effective evolutionary practice), those people having common ideas, and those common ideas being pushed through collective bargaining. So, the legitimacy of those common ideas can take the backseat to the fact that they are common ideas. This is often seen in victimhood narratives wherein the group has a story about how they were wronged (e.g. Holocaust, African American slavery), thus it excuses them doing whatever to reclaim what is theirs, or at least rebalance the previous injustice. The legitimacy of the victimhood narrative is rarely questioned, especially by those who benefit from it.
However, tribalism does have the positive benefits of preventing freeloading and keeping people loyal to a group. These positive benefits need to be accounted for because freeloading is always a net-negative to a group (i.e. people taking resources without giving back) and keeping people loyal to a tribe means they'll do pro-social things for it, often without cost.
A possible solution to tribalism is to reengineer the human mind to be algorithmic.
This could first involve usage of CRISPR gene-editing to cut parts of the brain which generate tribalism, Genome Editing with CRISPR-Cas9 - YouTube (although CRISPR currently has limitations regarding the human brain (even a single cut can create toxicity in the braincell, causing it to die) CRISPR and the brain: how gene editing benefits neuroscience | IDT (idtdna.com) . Whilst the genetic sources of tribalism have yet to be discovered (AFAIK), cluster analysis and eventually gene isolation should advance sufficiently to identify the genes that contribute to tribalism Bioinformatics: Finding Genes (genome.gov) . If cluster analysis is the method used, comparing DNA profiles of differing people to determine their genetic makeups, and then comparing that to their attitudes towards tribalism, could be used to pinpoint the sets of genes generating tribalism.
The second part is to make the human brain more algorithmic, in order for reason to be at the forefront of decisions. Making rational decisions could be more emotionally weighted, so as to give the emotional impact of tribalism without the shortcomings of it (so that it can compete in a harsh environment, or perhaps the algorithmic mentality would be enough??). Putting reason at the forefront of human decision should drastically improve efficiency of societies. It could also make other inefficiencies, like motivation, obsolete. Theoretically, this could be done by improving overall intelligence ('g' factor) in humans (maybe through manual insertion of genes which are associated with higher intelligence Genetics of Intelligence - I.Q and Human Intelligence (human-intelligence.org) ) and expanding the human brain's memory capacity to overcome memory-shortage issues like Dunbar's number Dunbar's number - Wikipedia , in conjunction with the removal of tribalism. It's possible that other emotional shortcomings will need to be removed, too. This is far more theoretical than the first part.
This reengineering of the human brain is fundamentally different from the historical issue of science, atheism and reason being held up as the solution to problems, because humans are being physically changed to delete/modify the evolutionary baggage, rather than people thinking they can simply think above those shortcomings (i.e. thinking you're above tribalism when you're not, and making decisions based on thinking you are (PDF) Tribalism Is Human Nature (researchgate.net)).
Of course, keep in mind that I'm not a neurologist or geneticist. But these are some theoretical arguments that I've tried to make a bit more practical, in order to move past human tribalism.