If a product is good, we shouldn't judge it on its inability to continuously add new features.

Not sure I understand that. If I want feature X, and <app> doesn't have feature X, then that's a negative. Doesn't mean <app> is not "good", but it's certainly criteria by which to judge it.

If <app> is good, then its nice to focus sometimes on the fact that its good, rather than on what it doesn't have. Its similar to judging a Honda for not being a Maserati that gets 200 miles per gallon.

Sure, but whether or not I decide to stick with <app> depends on many factors. One of those is "does it do everything I need/want." If the answer is no, then that factors into my decision.

No app is gonna rub your back at night or hold your hand on your deathbed (well, yet, google is probably working on something), so "everything" isn't... really a thing.

Who said "everything"? You're just making up arguments I never made. Can we agree that an app with zero features is useless? Assuming we can, the utility of an application is a function of which features it has and how well those features are implemented. So the features are important, and how they're implemented is also important. I don't understand why that's (apparently) controversial.

Arguably, though, an app is more than just the sum of its feature list. Also, even if a feature is desirable when considered in a vacuum, that doesn't necessarily imply that adding it will improve the product.

Somewhere in between $very_long_ago and $long_ago, I stopped using a well-known personal finance app because I grew dissatisfied with it overall. It gained all sorts of genuinely useful capabilities over that time period, and didn't shed any, but still managed to become less functional (for my purposes) at the same time. I ended up replacing it with a competitor whose feature list was much, much, much smaller.

I just wish their imports were a bit more reliable. Although some of that comes down to the fact they use 3rd parties that scrape the screen of your bank, so any time something changes, it breaks the imports. (it would be great to get an API standard, with proper authentication, etc)

Also, there is a plugin that gives you more reports: https://www.toolkitforynab.com/

I love YNAB but definitely not for the reports. I've been downloading my data and analyzing it in Rstudio whenever I want prettier charts.

Agreed. Their reports are horrible and for all practical purposes useless. It has a few other warts and areas of improvement, which bother me, but none as much as the lack of reporting.