Say we're offering our opponent 1:1 odds on a river jam and we have 8 combinations of value hands in our range for this line. Does that mean we should select the best 8 combinations of bluffs we have in terms of card removal in this spot to jam with, assuming we're trying to play unexploitably?

Reads aside, I'm just wondering if there are any theoretical reasons to deviate from the 1:1 ratio of value to bluffs, or from selecting only the best combinations in terms of card removal for our bluffs

How? Did you just make an overbet of infinity * pot? Even Libratus hasn't worked out how to do that.

Dammit. I f'ing thought about that mistake and still made it. Let's say 2:1 on pot odds and 12:6 on combos

I select my river bluffs from the bottom of my range. I think bluffing with hands that can win showdown is exploitive. I've won pots with as little as seven high when my lone opponent chooses to give up with junk. I think this strategy performs better vs human opponents. However, I do use blockers to bluff the river in very narrow range spots, but I usually play very shorthanded so I don't come upon these spots often.

In that case, I don't think you can say "Always bluff with blockers", or "Always bet the absolute bottom of your range". The card removal effect varies from spot to spot.

If you look at what solvers suggest, there's a lot of mixing to start with (so it's not exactly 6 combos that bluff at 100% frequency, but more like 2 combos that bluff 100% and 20 combos that bluff in the region of 20% frequency, or 10 combos that bluff around 40% of the time, such that the overall bluffing frequency is "equivalent" to just betting 6 combos at 100%) and the importance of blockers varies from situation to situation. Sometimes, you're literally better off betting the nut low and checking all your mid-strength hands (that might win or chop very occasionally) and sometimes you have some combos where the blocking effect is so powerful that they make for better bluffs than total air. e.g. in a spot where 65o is the bottom of your range, and you want to bluff with 18 combos, you might only bluff with combos of 65o and 76o and 87o that contain a flush blocker (6 combos of each), as opposed to bluffing all the 65o combos (12), half of the 76o (6), but none of 87o.
FWIW, I think it's often very hard for a human to correctly assess the blocking effects when considering a bluff or a hero-call, particularly when ranges are quite wide. It's easy to say "I have the nut flush blocker" when you have the ace, but the way lower cards interact with wider ranges is super-complicated, because it's not as if villain's range is purely nuts or air. He usually has many hands where blocking the 11th or 12th nuts might have some small significance.
Since card removal is so confusing, and you're rarely making a massive mistake (in 6-max NLH specifically) if you just bluff with the absolute worst hands in your range, you probably don't need to think too much about card removal effects except when you've narrowed your opponent's range considerably. e.g. If villain only calls with sets and straights, use hands that block those for your bluffs. If he calls with all kinds of pairs and two pairs, such that blocking effects are incalculable, bluff your 5-high like a boss.

Great post

Thanks Arty. My big takeaway from this is that spots where opponents have weaker/wider calling ranges lean toward sdv considerations and spots where opponents have stronger/narrower calling ranges lean toward blocking considerations. That makes sense b/c like you said blocking particular hands only has a significant and apparent effect against a narrow range, plus our sdv is less of a consideration against a strong/narrow range. I can see it only getting more complicated when we have to choose between calling with sdv and raising as a bluff b/c of blocking considerations.