The question of Assad's fate in light of the rise of ISIL and their threat to Western homelands is up for debate:
The tide of global rage against the Islamic State group lends greater urgency to ending the jihadis' ability to operate at will from a base in war-torn Syria. That momentum could also force a reevaluation of what to do about President Bashar Assad and puts a renewed focus on the position of his key patrons, Russia and Iran.
I find it horrifying that the rise of ISIL is making Assad seem like a lesser of two evils that justifies supporting him, as Russia and Iran have long wanted.
When I wondered whether the threat of loose chemical weapons and endless death and destruction in Syria was too great a problem to be countered by overthrowing Assad to get somebody better to control Syria, I wondered if it would be better to think of a post-Syria Assad rather than a post-Assad Syria:
The Assad regime and the rebels seem to have mutually exclusive goals that will require many more dead Syrians to resolve[.] ...
But the impasse is solvable, I think, if we stop thinking of Syria as a unified objective. If we move beyond that, perhaps both sides can get what they want--if the light is just right. ...
We'd need to give the Alawites and their Christian allies a province they demographically dominate on the coast and mountains of western Syria; the Kurds would get a province in the northeast; and the Sunni Arabs would get the rest.
Damascus would be a federal enclave.
Power would be devolved to the provinces, including foreign affairs excepting some symbolic aspects reserved to the mostly nominal federal government. Assad could become president of the Alawite province and carry on as he has.
The Kurds and Sunni Arabs would get their own provinces. Perhaps the Sunni Arabs get multiple provinces based on existing provinces or groupings of them. The various rebel factions could be bought off with control of these provinces.
At the federal level, the rebels would dominate, with a rebel in the position of prime minister representing legal "Syria" at the UN. Powers might be restricted to the federal zone and certain legislative areas not reserved to the provinces, such as allocating oil revenue and tariff revenue to the various provinces.
So we and the rebels get rid of Assad as the ruler of a relatively strong Syria (he will be demoted to just another despot who can't do too much harm to us with his resources); the rebels eject Assad from the presidency of Syria and take charge of the national government as the dominant groups, and control their own people's lives in the provinces; WMD assets outside of the Alawite area are secured (I assume Assad will want to quietly keep some), perhaps with an international force to do the job; Lebanon gets to edge away from resuming the civil war; Israel gets a weakened threat to their north (although the Golan border could get hotter with terror threats, the conventional threat will be much lower for a long time); Turkey and Jordan get to send the refugees back to Syria; Iran and Russia retain a pocket of influence in a rump Assad realm to retain access to Hezbollah or naval and intelligence bases, respectively; the Kurds get the hope of an independent state in fact if not legally; tens of thousands of further deaths might be prevented; and the UN gets to feel that it achieved something through diplomacy.
Now that jihadis and ISIL have emerged (and the migrant threat to Europe), this threat could be the threat that means we have to compromise on Assad--but without fully siding with Assad and the Alawites as the lawful rulers of all of Syria.
And there is precedent for splitting up Syria--France (quoting this author):
After expelling the British-installed Hashemite ruler of Damascus in 1920, France created five separate Levantine states based on the old Ottoman vilayets (“provinces”): Greater Lebanon, an Alawite mountain state, a Druze mountain state, the State of Aleppo, and the State of Damascus. Concerned that a rising Germany was making inroads into its colonies, however, France acquiesced to a unified Syria in 1936, ending the short-lived experiment.
Add in something for the Kurds (as the best fighters they can't really be screwed in a settlement), perhaps the Christians, and splitting up the Sunni Arabs to dilute their power a bit in a federal system, and we might be able to mobilize Syrian efforts against ISIL and al Qaeda without ratifying Assad's control of Syria.
Without the need to be brutal to rule all of Syria, perhaps the Alawites themselves will, over time, turn against the Assad clan for the bloodshed they endured in the civil war.
If we pretend Syria is a unified state, we have the insoluble and mutually exclusive (at least in the short run) problems of getting rid of the butcher Assad and getting rid of the jihadi butchers.
I can wish we had supported non-jihadi rebels years ago to make them the strong horse of the rebellion rather than the jihadis. I could wish we had spent the last year building up non-jihadi rebels after we decided to fight ISIL. But neither happened. I've long thought that we could afford to risk putting the defeat of jihadis second after getting rid of Assad.
If defeating ISIL takes precedence over defeating Assad because we don't want to risk ISIL marching on Damascus before we can take down ISIL, then we have to consider how we can defeat ISIL first while demoting Assad so that he can be defeated later.