In the interest of science, I cleared out a chunk in my current pack and set a cube of various sizes filled with Carpenter's Blocks (the Barrier, or Fence).
I set the test within a single chunk. For me, a cube measuring 16x16x29 (length, width, height) - or 7,424 blocks - just barely started to register on my FPS. However, adding any further layers of blocks instantly dropped me from 60 fps to around 2 fps.
However, this was just assuming I did nothing other than walk around. Placing or removing any blocks whatsoever anywhere near that chunk resulted in a marked delay of a couple seconds per block altered.
For all practical purposes, lag actually starts appearing for block placement/removal at a much lower number - again on my system, I started noticing minor placement lag after only two layers (512 carpenter's barriers) were present.
I then changed out all of the carpenter's barriers for regular carpenter's blocks (untextured), again at the 16x16x29 size, and suffered a slight drop in FPS as it did with the barriers. However, there was NO lag in the surrounding area when breaking or removing blocks.
In fact, I was able to fill the chunk to the build limit of 256 (for me this was 45,560 blocks, as I started at Y76 on a bit of a mountain.
I suffered no block lag, and the FPS still was only marginally affected. For those with morbid curiosity, my FPS (unlocked, so not tied to the monitor frame rate) without any blocks placed, is around 494 FPS. With the carpenter's blocks placed to the height limit, it dropped to about 482.
Locking my framerate down to 60, like I would assume a number of more normal people might do, the framerates went from 60 (duh) to about 57.
Changing from blocks to stairs (carpenter's versions) apart from some initial lag, again settled to around 482 (or 57 locked) FPS.
So there appeared to be some merit to
@TomeWyrm's advice regarding the barriers in particular. I have not noticed the other blocks being particularly terrible yet. Carpenter's Ladders had some initial lag when I originally placed them in, but it also settled down to insignificant levels within a few seconds.
I then tried this with Agricraft fences, Minecraft Vanilla fences, and Immersive Engineering fences. None of them produced any significant lag at all.
I was unable to test FMP as I had no easy way to edit in the blocks with the mods I had at hand.