The problem is that if LTV <= CAC -- let alone under the 3:1 ratio -- you aren't making any gross margin/profit.
LTV = the expected cashflows on a gross margin basis, for the life of the customer. If you get $100 @ 80% gross margin, you have $80 gross margin profit per year. If you lose 100% of your clients after 3 years, you expect to make $80+$80+$80 = $240 of profit. Ah but you also need to discount those future cashflows (let's use a discount rate of 10%). So you really are making $80 + ($80 x 0.9)+($80 x 0.9^2) = $217.
That's your total LTV. Your total profit from the client before ANY "operating expenses" like Sales and Marketing costs and salaries, Rent, Overheads, R&D costs and salaries, office supplies, professional services, etc.
So OK you have $217 of LTV. But you have to spend $250 of CAC (which is "total sales and marketing spending and salaries divided by total new clients added") then you are losing $33 per customer BEFORE all the money you're spending on everything else required to run the business.
Now, of course, reality isn't this simple. And it's not all VC bullshit. Maybe this LTV is dramatically underselling the value of having that client, because in Year 2, you'll have New Product Module ready to sell them for $75. And your cost to sell (CAC) that module to them is $1 - since you've already done the hard work of getting them as a client. So now you adjust your LTV calculation and it's not quite as bad.
Or... by having 100,000 customers, network effects mean that either the lifetime of a customer is actually 10 years (vs 3) or the network effects mean you can monetize something else or... lots of arguments you could make for "grabbing the real estate now".
So if VCs -- who aren't idiots even if some are morally questionable -- see a path to greater LTV, you can kind of get comfortable with a higher CAC.
BUT AGAIN - the rule of thumb is that 3:1 is healthy, since that provides gross margin profit from the get-go, and any other LTV expansion from upsells is gravy.