The first category is a regular memecoin, except the deployer vibe coded a game and gave it the same name as the shitcoin. You can't do anything with the coin in the app, you might not even earn the coin by using the app. Their association is only in name, not in function. It's a coin that has an app, it's app wif hat, it's most alt-L1's.

The second category of appcoin are attempts at sitting on right side of the IQ curve. These are apps where the coin is functionally integrated into the app--think usage rewards, a staking layer, voting, etc. Their association is in function--"serious projects". Contentcoins live in this category. Or $ALFA, when Alfafrens was popping. Of course, the risk of venturing to the right side of the curve is failing, falling and landing on the mid curve.

The coin-wif-app category only sees volume driven by speculators. It helps if there is a token indexing platform for better distribution, but most of it is word of mouth. Both want number go up technology.

With token launchers, the coins are the product, but the canonical appcoin is often still the biggest volume driver with the highest market cap. Examples are $LAUNCHCOIN, $CLANKER, $TERMINAL, which only have a nominal association with their respective apps.

I think we haven't quite figured out yet how to make the latter category successful, and I also don't think the former category is morally worse. The core point here is that tokens are pretty good value carriers and excellent media for converting attention into financial capital. The nature and extent to which they carry out this task is probably not halted by tech anymore. There are infinite mechanisms to choose from and countless levers to pull on (AMM models, token supply, liquidity games, incentives), all in service of materializing value. More evidence points to coin-wif-app being a better recipe for success, but both are directionally correct. Just build things and experiment and just try what other people build.