The problem of Bf4 is that White has no great answer to ...dxc4 (with the threat of ...b5 defending the pawn).
If White goes a4, now ...Nd5 forces you to move the bishop again as you don't really want to trade knights and fix Black's pawn structure.
If you go e4 to stop the knight from going to d5, now you realize why the bishop would have been much better on g5. Think of a line like Bg5 dxc4 e4 b5 e5 (exploiting the pin on the knight). Now Black must go ...h6 Bh4 g5 Nxg5 hxg5 Bxg5 and White has a great position. But with the bishop on f4, Black can answer e4 with a direct ...Nd5.
That being said this idea for Black isn't that easy to find. While on the Lichess masters database Bf4 is hardly ever played and Black has a great winrate, if you go to the all games database, Bf4 is played quite often (only less than Bg5 and e3) with an about average winrate. Black only found the right reply 8% of the time. If you play around with the rating filters you'll see that ...dxc4 gets increasingly rare as you remove the higher end of the ladder.
So yeah I'd say it's cool to avoid Bf4 now that you know why it's bad but at beginner level there are definitely far worse mistakes and you should focus on those. I definitely don't think playing Bd2 is any better. I'd either go for Bg5 or one of the more solid plan based around e3 and an eventual e4 (prepared with a bishop on d3, a rook on e1, maybe a queen on e2 and so on).
By playing around with move orders you can see that Black's response to Bf4 isn't really a problem with this specific move order, but it seems that the idea exists almost any time White puts a bishop on f4.
As a final note, if you learn from the inaccuracies you play in your own games you'll soon know all the opening theory you need to know.