There are different benefits to each model.
Full game developed and released, completed: this is, I find, the most satisfying way for a game to be released. I know I can start the game and (assuming I do well enough at it or find it fun enough to keep going) actually complete the fucking thing in one playthrough, the story is more likely to have a satisfying conclusion, and I generally know much better exactly what I'm getting than the other development types where things can change drastically through development. The danger in this sort of thing, I find, is too much hype. Start advertising early and people just demand that the game be finished before it's ready, so I'd recommend going dark until you're near release. But really, this is what I most would want to see. Hell, if it turns out that Steam isn't too restrictive on the content of the games it allows on its platform (incest?), I'd much rather Patreon close up shop and people move their shit over to Valve's store and start releasing full porn games again.
Big updates: this style tends to work for games with big modular bits. As you mentioned, Summertime Saga comes to mind: each girl/set of girls can have their own story, come with minigames, and is sort of a small self-contained game in and of itself within the larger game. I imagine (not being a game dev myself) that this can be pretty hard to manage logistically speaking, but some devs can definitely pull it off. However, this does have the disadvantage again of the continual hype train: most months there won't be any releases, and that can frustrate players, but burnout on the part of the players shouldn't be a big issue. The larger updates should allow players to replay the game infrequently enough that things still feel fresh if it's their sort of game.
Lots of small updates: the Patreon way. This is the method for people who know, in advance, that they have the businesslike midset of getting through a lump of work every month, no matter what. More often than not, developers start to give excuses and slow down and just generally get tired of working a few updates in, sometimes even if they're being paid a buttload of money to get content out, though this usually happens because they didn't get the traction they were looking for with their audience. I don't know how many games you've developed before, but personal excitement about a project only seems to get most devs through the first 2-3 updates. After that, it turns into a job where the dev sets their own hours, and that tends to work about as well as you'd expect. From the player perspective, this also tends to start out exciting, but as the updates inevitably become incremental, the players tend to get really burned out on replaying the same thing over and over with relatively little change in content. Occasionally you find your niche with players who obsess over your work and are cool with replaying every release no matter how small or simply bug-fix-specific, but I wouldn't depend on it. On a personal note, I don't think there's any game that does monthly releases that I bother replaying every update for anymore. I generally check in on the games I like about every 3 months.
A side note on replayability: some devs confuse replayability with having lots of options. Lots of options means that players can choose to play different ways, but doesn't always imply that every choice is viable or even that every player wants to see the results of every choice. For example, a VN might give me the choice between banging the hot busty milf or the flat skinny jailbait girl. I'm always (ALWAYS) going to pick the hot milf, and somebody else is always going to pick the flat skinny jailbait girl. So that means that while you've doubled your development time on that section and perhaps appealed to a wider audience, you haven't really improved replayability for either player. This isn't a big deal in complete games, but seems like it has to potential to really frustrate devs who allow for a lot of choice in frequently updated games like SuperPowered where most players just won't bother doing what it takes to see all the new scenes every month when each new choice requires an entire (increasingly long) playthrough to get to. On the other hand, a well-written linear story can be worth reading/playing again just because it's that good. See: Akabur's best games, most specifically Princess Trainer Gold. While there's a lot of variation in scenes, you pretty much see everything there is to see in one playthrough, yet I've gone back to play that one several times over, and that game was released as basically two big updates (after beginning life as a longer side story in an earlier game).