Damage effects of CBU-99s with FMU-140 are hit or miss...literally.
I've been testing them for some time, and the best results I have are with CBU-99 with Mk 339 at low altitude.
For FMU-140, as per ED's instructions on September 7th changelog, for DCS 2.9.20.15010, under the Weapons section it reads:
Mk-20/CBU-99. Improved bombsight accuracy. Please note that using FMU-140 in low-altitude horizontal flight bombing runs will lead to poor accuracy - in these cases Mk 339 fuze must be used. FMU-140 should provide acceptable accuracy in dive bombing runs and, in specific conditions, in horizontal flight bombing runs (function altitude set to ca. 1/3 of release altitude and specific airspeed at release, e.g. function altitude - 3 kft, release altitude - 10 kft, TAS - 580 kts)
Mk-20/CBU-99. Fixed Mk 339 fuze function delay being counted down after arming instead of after release
In my testing, FMU-140 works best by flying level,dropping at 10,000 feet, with an airburst of 3,000 feet AGL, using AUTO mode.
This way the bomblets have enough time to arm and spread. The spreadsheet below gives you an idea of various profiles for dropping these cluster munitions. (If you are asking yourself why would I spend too much time doing this and testing, the answer is: I'm a stubborn nerd).
So, the end game is that you must allow time for the bombs to separate from the aircraft, open and, most important, arm.
However, the bomblets do cause damage only when they directly hit the targets. Funnily enough, the result is that -99s are most effective against armored than against light targets. I blame that on the lack of a proper fragmentation model, but ED said they are working on it. From the July 23rd changelog:
Weapons. Work in progress fixes for the following bombs explosive mass: M117, Mk-80 series, and WWII AN-M GP series. This also applies to Guided Bomb Units (GBU) that use the Mk-80 series of warheads. This is connected to the changes for SAMP bombs that we did in the previous patch and will be evaluated and finalized in future patches. Most rocket warhead families (at least Hydra/Zuni/FFAR, S-5/8/13/25, and SNEB-68) are already set up with parameters that match our available sources. Actual effect on targets is subject to change along with implementation of the fragmentation model. (We missed mentioning this in June 18’s update).
Try to hit the same targets with Mark 82s, and better yet, Mark 84s. You don't need to hit light targets directly, they'll catch fire from the frag effect.
P.S.: I forgot to mention that the -99s are falling short, either in AUTO or CCIP mode.
CBU-99 Height of Function.xlsx