The Legendary supertype has always been a bit of an odd duck in Magic. For fluff purposes it denotes a character, object or location that exists in whichever story the current block is telling, but for gameplay purposes it’s usually just a drawback with (often) no matching advantage. Exactly how much of a drawback depends on the era in the game’s history you’re talking about:
Original rule: once a legendary permanent is in play, no permanent with the same name may come into play.
Kamigawa rule: if two legendary permanents share a name, both are destroyed.
Current rule: if you put a legendary permanent into play and you (not your opponent) already have a permanent of the same name, you choose which one is destroyed while the other lives.
NB: “destroyed” here actually means “is put into the graveyard as a state-based effect.” Magic!
You can see there is usually not much benefit to being legendary, and it can be kind of a pain in the arse. Things changed when Wizards officially endorsed the Commander format, which allows a single legendary creature to be used as an always available resource. Any legendary creatures printed after Commander was adopted should at least have had some thought put into how they would perform as commanders. In this light, Grenzo doesn’t make a whole lot of sense.
The first thing to notice is that Grenzo was a card from Conspiracy, a set intended to be drafted rather than be used in constructed decks, which is fine. There are several other legends from conspiracy that makes sense in both a drafted format and as commanders, but Grenzo kinda sticks out.
In terms of base usefulness he is a vanilla creature of whatever power and toughness you can buy, which is decent in black and red but not spectacular. His activated ability, on the other hand, seems to do a few things but none of them particularly well:
1. Inefficiently mill yourself, to power reanimation or other recursion effects.
2. Maybe, in extremely lucky circumstances, get you a free creature.
3. Waste your mana and time.
In a draft environment where you have limited control over what creatures you will have in your deck there is a good chance that most of the time Grenzo’s power will not match up with the power of the creature (assuming it is a creature are not land or something) on the bottom of your library. It’s basically a gambling mechanic with odds stacked against you unless you draft in a very particular way. But hey, he’s far from the worst card to ever see play in limited.
Let’s look at him as a commander, and this is where his design really begins to make no sense.
If you’re running Grenzo is a commander (full disclosure: I do!) Then what you’re most interested in is getting the greatest possible return on investment for the two mana per activation you feed him. This means you want, as often as possible, the card on the bottom of your library to be a creature with power less than or equal to Grenzo’s. This is directly at odds with the way wizards seem to want him to be played based on his design.
Consider the X in his casting cost. Black and red have cards that put +1/+1 counters on themselves, but very few cards that put +1/+1 counters on other things. You can bolster his power with equipment and the like, but most of the time the size that Grenzo is when you cast him is the size you are going to have to work with for the duration of his life. But that’s okay, black and red aren’t amazing at mana acceleration but you’ve got your Dark Rituals and your Seething Songs and your broken lands and so on. You can totally build in a big enough base of mana acceleration and power pump to get Grenzo up to a big size. That’s half the deck right there, fill the other half with huge creatures (probably around the four-to-six power mark because that’s where Dragons tend to sit at and red has lots and lots of dragons) and we’re off to the races.
Except when you actually try to play this deck you will very quickly hit a problem: all the stuff you put in there to make Grenzo big is the same stuff that is going to be inconveniently on the bottom of your library half the time when you’re trying to get out creatures. The biggest offender here is land, most commander decks will run between 36 and 40 land depending on their mana curve. Add in 10 to 15 things to help out with Grenzo’s size and you can very easily hit a ratio of less than 50% creatures. Those are terrible odds.
This is the core of his design weirdness: he demands an infrastructure to most effectively leverage his ability but the ability itself punishes you for wasting card slots on doing that.
On the other hand, if you run some suicidally small number of lands (like 30) and just make every other card in your deck a one or two power creature then Grenzo works really well! In the months I’ve been playing him I have cast him with a nonzero X cost exactly twice. The first time I immediately lost him to a Cytoplast Manipulator; the second time I won because I needed a three power Grenzo to flip a Godo (which I knew was there), because I needed Thornbite Staff.
Thornbite Staff is ridculous.