I tried an experiment twice, one worked out well and one failed miserably. I recommend people try it, especially if you're just looking for a fun, different approach to the game (which I try to do every game.)
When starting a game go to "advanced game setup" (BTW, it's pretty common knowledge but always do this - if you aren't there are many game options you don't know about) Select "allow policy saving." The primary purpose of the option is to allow you to delay selecting a policy until a later turn, and the most applicable use is to delay policy selection until a tree unlocks. Most consider this abuse and cheating because it allows you to go straight from tradition to two or three policies in rationalism; fix this by simply not abusing the ability with the exception of the proposed experiment.
Do not select any policies until you enter the classical era, and then have Patronage be your first policy policy tree. Play the game as usual from there.
For the record, it's not an effective strategy, rather an interesting and different approach to the game. Tradition starts , Liberty starts, and even Piety starts are much more effective strategies, and even Honor starts give you some combat bonuses and GG's to help keep you alive long enough to make up for the error of selecting honor.
It is REALLY HARD to scrap through the ancient era without the assistance that the social policy bonuses give you. I chose Babylon both times to compensate for the extra difficulty.
By doing so, you can appreciate the full value of the patronage tree and particularly the finisher. The problem that I and most of us have is that we use Patronage as a filler tree after completing Tradition (or liberty) and before we can open Rationalism. As such, we then finish Rationalism as well as get the goodies from the ideology before we go back to Patronage, which means far fewer in which we can be gifted great people, and the game is usually decided by then. When you try the experiment, you solve this problem but encounter another: early game you don't have the infrastructure to maintain alliances.
The first game I tried was a failure. I didn't get many quests which allowed for fewer alliances so there were fewer great people. I got a bunch of artist-class great people that I had no slots for and would be wasteful to bulb (GM's with no tourism-per-turn, GW's writing treatises for 200 culture.) Got a khan but had no intention of warring with anyone due to my status. Ended up rage-quitting.
Then I tried another time, which admittedly was on a much better map (salt start.) What made the difference was domino-quests - multiple CS's targeting the same encampment, eliminating that encampment gave a resource which triggered two more quests, alliances giving vision which introduced another CS, etc. IIRC, that game I had 7 or 8 academies before I was in the industrial era (partly due to Babylonian UA but at least two or three gifted great scientists.) Two wonders that were from Great Engineer bulbs, ended up being a successful turn 260ish diplomatic victory.
Again, it's not an effective strategy compared to other options, but a fun and different approach to the game. Also very luck-based, akin to Spain games. But it allows you to utilize a component that you rarely get to.