This post speaks to the business of picking a driver suitable for a given LED. Apologies for busted links (it’s ~10 year old content…) but the general ideas remain valid. Note that if you’re matching an LED to a driver, you’d need to work in a somewhat different direction.
That said…
These LED ratings would mean that if a person shoves 700mA of current through the device, about 1000lm of light will come out, and one will need about 14V available in order to get the specified 700mA to flow.
Note that 0.7A*14V=9.8 watts of power, roughly half of which would have ended up getting turned into heat instead of light.
A driver specification like that means a device that is designed to shove 350mA through whatever is connected to it, and not complain so long as the voltage required to do that ends up somewhere between 10 and 35V.
In other words, while the old LEDs might have been rated for 700mA, they would have only been operated at 350mA. Lumen output (i.e. total quantity of light) is pretty much proportional to drive current, so an LED rated for 1000lm @ 700mA would only be delivering ~500lm @ 350mA.
Also, it would only be giving off half the amount of heat @ 350mA that it would if run @ 700mA. That’s important, because LEDs very much do not like heat/temperature.
If we look at the specs for the CDM-9-3018-90-36-DW01, we find a test current figure as well as a max current figure. The flux rating and forward voltage apply at the specified test current and temperature.
If you tried to run that LED at 350mA instead of 300, the forward voltage is going to be a tad higher and pushing slightly beyond the stated limits of your driver. (The relationship between voltage and current is not proportional, which is why we have to mess around with constant-current drivers…)
Under those conditions, one would anticipate slightly north of 1000lm out of the device, which is roughly twice the light one would have actually been getting out of the old setup. Additionally, the total power burn would be roughly 0.35A*36V = 12.6W, roughly 2.5x more than original.
What I’m getting at here is that the CDM-6-3018-90-18-DW02 originally mentioned is probably the better option of those discussed so far. Light output should be roughly comparable, and while the power burn/heat output would be slightly higher than the original, it’s 'aint no 2.5x higher.
Part of the reason this CoB is less efficient than some is that it does a bunch of fancy gizzyfritzing to end up with a light color that shifts more red/yellow as it dims, like an old-fashioned incandescent did. It also goes to some pains to produce light that doesn’t do odd things from a color perception perspective. (CRI=Color Rendering Index: higher=less funny business, lower usually=more lumens/watt, all else equal)
My suggestion to you would be to clarify what’s tolerable from a mechanical mounting standpoint, and where you’d like to go from a color temperature standpoint. Higher color temperature (CCT) are called “cool” and are more blue, lower CCTs are called “warm”, are more yellow, and deliver less light/watt in general. Pretty much a preference thing.
There’s actually a few hundred part numbers in stock that would seem in-ballpark from an electrical standpoint once a person figures for the whole operating-point business, once a person figures for the whole rated-for versus operated-at business…