It's probably worth separating out the mechanical and UI aspects of your question. You could remove mechanical color (in the narrow sense that you're asking about) and keep color cues on the card frame - and in fact, keeping color cues would be just as important, because references to mechanical color come up immeasurably less often than wanting to quickly know what kind of mana something costs in limited. (And note that, even if there were no cards that referred to mechanical color, and narrow mechanical color wasn't defined in the comprehensive rules, everyone would still talk about what color things were.) Likewise, you could remove color cues from the card frame and keep mechanical color.
My view on these is:
1) Mechanical color isn't essential, but will appear intuitively obvious in most cases, so the cognitive burden is much less than with many mechanical hooks. I think the design space is worth it, but that's a subjective decision, and either way one that won't affect play a great deal.
2) Card frame cues about what color something is - including in the colloquial "what mana do I use to cast this" sense - are indispensable. It's useful for the pro player who needs to shave off microseconds of filtering out irrelevant information when drafting, it's useful for the newbie playing Magic League whose knowledge of deckbuilding is "pick two colors and a bit of a third and make your deck out of those," it's useful for people sorting the cards they drafted into their deck boxes, it's useful useful useful.
Now, it is true that cards that are a different mechanical color from the color required to cast them have a frame that indicates the mechanical color. But these are rare, and most of the time the difference is one implemented for flavor reasons. It's neat to see cool, collected Dr. Jeckle become the raging read Mr. Hyde, or the sweet little girl become the evil demon - but that's also something where the presence of color on the frame helps tell the story.
Miscellaneous legibility improvements. Since I don't have a lot of experience teaching the game, and they don't effect much other than that, I don't feel like I can say much here.
Tagging tech. This allows us to key things to tags to a much greater degree than before, or at least with easier templating - obviously the heavy lifting of how this works out in practice is a matter of set rather than rules design. Probably a lot hangs on choices of how many evergreen tags are lying around and what their mechanical significance tends to be. I like this insofar as anything that opens up design space is good.
Rarity changes, like where to place nonbasic lands in packs. Reduces costs of entry, which is good for Magic the game if not Magic the revenue stream, as you point out.
Haste and tapping to block. These were adopted for legibility reasons, but they change how players evaluate combat math a bit, which is probably worth considering.
No colorless artifacts. Etherium Sage above brought up some of the consequences of this.
Conditions. Huge change. Opens up a lot of design space, reduces variance, precise effects depends a lot on card- and set-level design.
Rarity-based limits. Increases variance, reduces fixed costs of entry. I wonder if global rarity rules ("you can only have N $rarity cards total") might be useful in addition to local rarity rules ("you can only have Z of any given $rarity card"). One problem with solely the latter is that it hampers goofy Johnny jank, though maybe some of the role played by build-around-me rares can be played by Conditions.
Separate Land and Spell decks. This is by far the biggest change, though a lot of the relevant considerations in moving to a Duel Masters-style system have been discussed by the community elsewhere. Reduces bad variance. As a technical consideration, I think you definitely want to have separate minima for the two decks ("At least 30 cards in your spell deck and 20 cards in your land deck,") because ratios no longer matter - there's no longer any reason for Sligh to pack a lower percentage of lands than Draw-Go; that's moved onto the latter exploring more often and drawing less - and by your current RAW, if I'm reading it correctly, a player can force their spell deck's variance arbitrarily low, which I don't think is your intention.
I wonder if Planechase might make a better model for Conditions than Conspiracies? That is, you can have a couple of different Planes, all of which support your deck's theme, but you don't have a ton of control over which is active at any given time. Plus it's flavorful! (Or "resonant," if you must.) This opens up some design space in terms of mechanically affecting what Plane you're on (one of the tragedies of Planechase itself - no new cards that interacted with it) and adds in some variance - a more interesting kind than that getting sucked out by the separate Land/Spell decks.
(Some additional interesting variance might be keyed to tags on lands, since what lands you pull are random.)
Re: legibility/logistics, I'm not sure the increased granularity of 30 rather than 20 life is worth the end of being able to track life in most games with a d20. This is of course a minor thing, but so is the increased granularity.
Anyway, looking forward to where you go with this! Most of my substantive comments have been on the critical side, but that's mostly just an artifact of there being more to say where one is critical.
My view on these is:
1) Mechanical color isn't essential, but will appear intuitively obvious in most cases, so the cognitive burden is much less than with many mechanical hooks. I think the design space is worth it, but that's a subjective decision, and either way one that won't affect play a great deal.
2) Card frame cues about what color something is - including in the colloquial "what mana do I use to cast this" sense - are indispensable. It's useful for the pro player who needs to shave off microseconds of filtering out irrelevant information when drafting, it's useful for the newbie playing Magic League whose knowledge of deckbuilding is "pick two colors and a bit of a third and make your deck out of those," it's useful for people sorting the cards they drafted into their deck boxes, it's useful useful useful.
Now, it is true that cards that are a different mechanical color from the color required to cast them have a frame that indicates the mechanical color. But these are rare, and most of the time the difference is one implemented for flavor reasons. It's neat to see cool, collected Dr. Jeckle become the raging read Mr. Hyde, or the sweet little girl become the evil demon - but that's also something where the presence of color on the frame helps tell the story.
I wonder if Planechase might make a better model for Conditions than Conspiracies? That is, you can have a couple of different Planes, all of which support your deck's theme, but you don't have a ton of control over which is active at any given time. Plus it's flavorful! (Or "resonant," if you must.) This opens up some design space in terms of mechanically affecting what Plane you're on (one of the tragedies of Planechase itself - no new cards that interacted with it) and adds in some variance - a more interesting kind than that getting sucked out by the separate Land/Spell decks.
(Some additional interesting variance might be keyed to tags on lands, since what lands you pull are random.)
Re: legibility/logistics, I'm not sure the increased granularity of 30 rather than 20 life is worth the end of being able to track life in most games with a d20. This is of course a minor thing, but so is the increased granularity.
Anyway, looking forward to where you go with this! Most of my substantive comments have been on the critical side, but that's mostly just an artifact of there being more to say where one is critical.