I'm afraid that I'll have to say some things you might not want to hear now, herw.
The first is that suggestion you always refer to. It's a half sentence in the Alchemy rules. The full sentence reads like this:
You can play Dominion using any number of Kingdom cards from Alchemy, but we especiallyrecommend playing with 3 - 5 Alchemy cards at once
It says right there that this is not something you have to do. In fact, most people don't do it. It is a hotfix Donald X included to make sure that in case the card he put together under a lot of time pressure by the publisher would still work fine. We all know that to buy a Potion card you need to get a Potion. What a lot of people are unaware of is that this means that, if you only ever want one Potion cost card, you would end up with a 50% dead card quota from getting it and Potion. That's awful. To remidy this, Donald X made sure that all Alchemy cards are cards you want to get multiples of, even if there was only one of them in the kingdom.
The suggestion in the rule book is a safety net. Assuming that this was something you could count on, treating it as a fundament of your designs, puts too much strain on it. As a result, you constantly underestimate the opportunity cost of Potion cards. You rely on the value on the official Potion cards, hoping that they would make yours worthwile; but frankly, several of the official Potion cards are horrible, and you are adding further weak cards. As a result, the net grows thinner and thinner, with the result that even if you'd set the game up to always have three Potion cards in the kingdom, now they are more likely to be three weak Potion cards.
As a general thing that I learned over the course of several years: Don't try to make cards with the main goal to fix something. Don't make a card to boost Scout, or harm Rebuild (those are things that I actually tried in the past). And don't make cards to boost Potions. Because if you focus on making cards Potion costs or to boost Potion cost more than you care about making good cards, you will start attaching Potion costs to cards that don't need or want them. You will be missing out on good card ideas by shoehorning them into something that doesn't work with it. In the end, the cards will end up generally worse because of this.
I have stated this before: Potion cards are very tricky to do right. Look at Alchemy, and how many cards of it are bad. Even Donald screwed up several of them. If you are a relatively inexperienced card designer, Potion cards are one of the worst places to start with. But if you must make a Potion card, at least try to apply the same constraints that Donald applied. Make cards attractive in multiples. Make them work well in kingdoms without other Potion cards. Perhaps you can learn from Donald's mistakes and make the cards not too slow. But don't lean on a safety net. Thats not what it's for and it can't carry that weight.