Conventional Weight Gain Wisdom Seems Off

There’s a lot of people “out there” that want to lose fat while gaining muscle. I don’t believe that works - or at least works so poorly compared to conventional bulk/cut it’s even worth considering. However, what about gaining muscle while keeping bodyfat steady? Can it be as effective as the alternative?

Here’s my logic based on an incomplete knowledge of nutrition - but bear with me:
First, take your resting metabolic rate (RMR - the amount of calories you burn sitting on your ass all day) - then add whatever calories you burn lifting. OK, great, so let’s say it’s 3500, then (ignoring macronutrient and food quality breakdown) if you eat 3500 calories, you’ll stay the same weight (I know this is oversimplified but it’s unimportant here).

So - conventional weightgain wisdom = eat (a lot) more than RMR+cal burned, gain a ton of muscle and some fat, get hyooge, cut, be beastly strong/huge + get girls.

But, why is this the best way? Let’s say you stay within this 3500 calories, but eat intelligently. You get your 1g/lb or thereabouts of protein (from good, whole, natural sources ideally), eat tons of fruits/veggies/healthy fats for the rest.

Maybe break it down like this for a 200 lb person: 200g protein (aka 800 calories), and the other 2700 calories, 1500-1800 of fruit/veggie/healthy fat + some low g.i. carbs, and then the remaining 900 calories use for strategically timed insulin spikes to load muscles with protein.

It seems that then you’re getting the best of everything. You’ve got ~800-1000 calories to play around with insulin spikes and stuff for filling your muscles with protein and glycogen, you’ve got tons of goodness from your fruits/veggies/healthyfats, and you’re avoiding over-the-top weightgain.

In one sentence - why is (even if it’s good, wholesome food) eating way over RMR+calories burned better and inducing muscle gain than eating RMR+cal intelligently?

As mentioned above, my knowledge of nutrition is lacking, and I may be missing some huge point here.

Thanks, sorry for a long read!