I find that in order to have a cube with interesting lands and decks with competitive manabases, drafting 45 cards just isn't enough. Here's my "back of the napkin" math on the subject:
A deck is going to have about 24 non-land cards. A decks with a competitive and interesting manabase will have at least 8 non-basic lands. That's 32 cards, so 13 drafted cards won't make the deck.
Of those 13 cards, 3 of them will be last picks. Because you didn't pick them, there is a high probability that the card won't fit your deck due to any number of reasons. Many of the cards that are last picked are going to be narrow cards where the narrow deck they tend to get played in isn't at the table. So lets say you get unlucky and none of your last picks are useful to you. We are down to 10 cards.
Of those ten cards, 3 will be next to last picks. The problem here is similar: with only two cards to pick from, there is a good chance that neither will be useful to you, but since you have a choice, its more likely one will. Let's say one of them makes your 40, so two of them don't. We are down to 8 cards.
Of those 8 cards, 3 of them will be 13th picks. Same problem, but with three cards, lets say two of them are useful. We are down to 7 cards.
Of those 7, at some point in the draft from 12th pick up you are going to see an occasional pack that just sucks for you and has nothing in it. Lets say its one pack. We are down to 6 cards.
Of those 6, you are probably going to need a couple of cards to sideboard into certain matchups, to let you sideboard out cards that are not helpful in certain matchups. Lets say you draft three cards for sideboarding. We are down to 3 cards.
During the course of the draft, you are going to take early picks that are speculative. However, the deck or color might get cut, the synergies might not work, etc. Let's say you speculate 3 times during the draft. That's all 45 cards.
And it leaves no margin of error whatsoever. Sometimes the packs break funny. Sometimes you need more lands. Sometimes you don't realize how the openness of things until pack 2. Things do wrong and there should be some cushion for players to maneuver. Obviously in a card game there is going to be luck, but a margin of error is nice because it lets good players work around bad luck and it lets bad players recover from their mistakes. The meta-goal is get a deck that is fun to play, so I don't see a point in the drafting process making that overly difficult. Better play is still going to lead to better decks which will still lead to wins, but at least people won't be miserable while they are losing because their pile has cards that don't fit or shit mana.
The two ways to do this are draft more cards by increasing the size of the packs(I think 51 is much closer to right amount of cards to be drafted when you are trying for ~32 playables as opposed to ~26 in a retail draft) or draft certain cards separately (Utility land draft or something similar).
Related to the recent posts, I ordered 8 Evolving wilds and I am going to throw them into my ULD in order to further support the upcoming landfall theme in my cube.