Starting to Sum Up

At Pacific Standard (a casualty of the collapse of niche publications), Seth Masket wrote about a major theme of this course:

For several years, a number of scholars (including me) have been making the case that American political parties are best thought of as loose networks of interest groups, candidates, donors, activists, and others, rather than hierarchies organized under the DNC or RNC. (You can see examples of such studies here, here, here, here, and here.) 

This observation is important for historical context.  When Madison wrote about parties, he was talking about what we would today call PIG -- specifically, legislative factions. There were no party organizations as we know them today.  Indeed, Americans tended to frown upon overt political campaigns. This exchange from Hamilton makes the point:

HAMILTON:  You've created quite a stir, sir
BURR:            I'm going door to door
HAMILTON:  You're openly campaigning?
BURR:            Sure.
HAMILTON:  That's new.

 Masket also made an observation that may relate to your essays:

The way we think of parties is vitally important to how we treat them. If we want to regulate parties or restrict their activities or even ban them, that's far easier to do if they're rigid hierarchies than if they're flexible networks. The latter can adapt to a great many impediments.

The same is true for proposals for party strategy.  If you argue that the Democratic Party should do X or the Republican Party should do Y, remember that there is no central shot-caller for either party.