I wish there was more dedication to model physics/background in school (maybe that's more of a grad level thing or maybe other schools do this). Understanding models is more than just being able to interpret the output. That is a small, tiny fraction of it. Ultimately though, forecasting has gotten more lazy. When I first joined the boards in the mid 2000's there was much more in-depth analysis and assessments of patterns, synoptics, etc. When you go around and read accounts of how meteorologists used to forecast back in the day and the techniques used...it all seemed very thorough.
Now when there is a severe weather threat in the Plains, everyone runs to supercell composite charts and significant tornado probs, anytime there is a winter weather threat, it's running to the 10:1 maps and using 10:1 maps for snow/sleet, even though it gets stated 8 million times why you can't do that, or running to the Kucheria snow maps because it paints out more snow than the 10:1.
The integration and explosion of social media and how irresponsibly alot of this information gets tossed out makes it seem like models are worse but they really aren't.