The posters at MTGSalvation tend to be a hidebound, traditionalist lot, so don't expect a lot of support for your heretical proposals here. I recall seeing similar suggestions in the past decried as "not real Magic"; I find such arguments wholly unconvincing. Seeing as no one has made them in this thread, though, I will not refute them here.
I've implemented your idea in my own cube by modifying the mana cost of a handful of cards to make their power level more in line with the rest of the cube. I put a tiny shiny sticker next to the card's name, so the players know to take a closer look at the mana cost during the draft. I've found that tweaking the mana cost (as opposed to altering the card's ability itself) tends to be least obtrusive way to adjust a card's power level. Once the card hits the table, its mana cost is rarely significant, so it won't matter if the players forget its new cost.
The biggest advantage of your proposal is that you can enable entirely new archetypes that don't have quite enough support in the current configuration of cards. There are some themes (particularly in older sets) that have been power-crept into obsolescence, but could provide for rich and interesting play experiences if they only had a couple more support cards at a higher power level.
As for your question about how to know how much to adjust the cost by, there's really no precise algorithm - you just have to go with your gut. Ask yourself if the card at its native cost would be either a) a crazy bomb or b) relegated to side boards for being too weak. Then adjust the mana cost up or down by 1 generic mana appropriately. I tend to be more willing to increase the cost of a card then decrease it. Lowering a card's cost improperly can result in blow-out games that leave players feeling bad; raising a card's cost will merely send it to the sideboard. After a few drafts, you can adjust the cost further (or undo the change). For example, I run The Abyss and Jace, the Mind Sculptor at 5 CMC. I run a handful of cards at lower CMC's too - e.g., Trail of Evidence. So far my players have appreciated the chance to play with cards like Trail that they wouldn't normally have because they were too weak.
I've implemented your idea in my own cube by modifying the mana cost of a handful of cards to make their power level more in line with the rest of the cube. I put a tiny shiny sticker next to the card's name, so the players know to take a closer look at the mana cost during the draft. I've found that tweaking the mana cost (as opposed to altering the card's ability itself) tends to be least obtrusive way to adjust a card's power level. Once the card hits the table, its mana cost is rarely significant, so it won't matter if the players forget its new cost.
The biggest advantage of your proposal is that you can enable entirely new archetypes that don't have quite enough support in the current configuration of cards. There are some themes (particularly in older sets) that have been power-crept into obsolescence, but could provide for rich and interesting play experiences if they only had a couple more support cards at a higher power level.
As for your question about how to know how much to adjust the cost by, there's really no precise algorithm - you just have to go with your gut. Ask yourself if the card at its native cost would be either a) a crazy bomb or b) relegated to side boards for being too weak. Then adjust the mana cost up or down by 1 generic mana appropriately. I tend to be more willing to increase the cost of a card then decrease it. Lowering a card's cost improperly can result in blow-out games that leave players feeling bad; raising a card's cost will merely send it to the sideboard. After a few drafts, you can adjust the cost further (or undo the change). For example, I run The Abyss and Jace, the Mind Sculptor at 5 CMC. I run a handful of cards at lower CMC's too - e.g., Trail of Evidence. So far my players have appreciated the chance to play with cards like Trail that they wouldn't normally have because they were too weak.