Yes, it’s not exactly the same square waveform which does change the readings… Only the raising current edge is the issue and that being so short, I think you can ignore it.
On my scope, I can see the waveform and I’ve found it’s close enough you can’t read the difference on an analog meter. This is using a 0 to 30mA range.
It works well enough to use it, at least with these machines.
Try it and see…
My understanding is that PWM modulation of average current is sufficient for protecting the tubes, because what kills them is insufficient heat flux inducing failures in CO scavenging.
I have not seen this anywhere… where did you find this?
Keeping the temperature under control and following your logic, if I have a 20mA maximum current limit, it’s ok if I run it at 40mA for half the time or … 60mA for 1/3 the time? … I don’t think so and I think you’d agree … Tubes are designed to operate at various current levels, unlike a ssl. These, K40’s are configured to be run by a system that drives ssl, not tubes.
No need to argue the 60mA/time is pretty ridiculous … When you set the current level like this with a dc voltage, when it lases it lases at 100% and it’s accumulative. 100% of the time these lase, they lase at 100% power… or whatever your dc voltage current limit is.
Most of these low cost tubes have at least a 2000hr expected life time (MTBF). You’ll notice that to get this life, you have to use it <40% power which is <12W… I use mine, mostly in the 40 - 70% range, but I cut in the high end. You end up, if over 70%, almost a 75% loss of tube life.
The Ruida will run other than 100% power continuously, not true with the standard grbl type boards for a ssl that the Chinese use. Most of them are copies of western world controllers, similar to the Arudino… software was probably hacked western world open source.
All information I’ve ever seen, including on this site and every where else is don’t over drive the tube and it’s always directed at the current flow. All information I have says current is the main cause of damage to a tube.
I agree, that heat, is an issue. Anytime you want something to work faster chemically, you generally heat it up for faster reactions. I think my tubes life was shortened substantially in Phoenix as we had a very long stretch of 53 days that the outside temperature was 110F or higher. My co2 machine lives in the garage … No a/c
Some of these people claim to have gotten extraordinary life from these K40 configuration and I have no reason to suspect otherwise.
Operating normally, the guys with all the degrees say it’s down to how well aligned the mirrors are in the amplification area, will dictate how long they last… This is way over my head, but thought I’d throw it in anyway…
I’ve notice more of the K40 types are configuring pwm to drive the IN terminal and using L or H for a laser enable. They shouldn’t lase if there is a 0 voltage driving the current limiting circuits of the lps.
If I configure my Ruida for an RF source, it operates like it’s driving a ssl, there is no laser enable in that mode. No pwm, no lase, same lps.
At least that’s my view… hope you’re not sleeping by now… 
It’s my interpretation of at least what I read in Sam’s Laser FAQ: CO2 and of what I’ve read here by @donkjr about how the laser power supplies work. I’m confident that I have read other resources as well.
Basically, two completely independent points:
This is also supported by the point of using an analog ammeter not a digital one; the inertia of the needle averages current to give you a reading that corresponds to effective power, whereas cheap digital ammeters can have bad aliasing that cause the reading to jump around. (This is different from a digital voltmeter sampling the PWM input that is nearly useless except to help identify a tube that is losing power.)
Doesn’t mean a potentiometer is a bad idea. But the rationale I’ve seen shared here on Maker Forums multiple times is that by setting the pot to 100% of the intended power, you can’t accidentally over-drive the tube from software, and you get to use all 256 (or whatever the precision of your hardware) levels in software.
Additionally, despite you saying “including on this site” I don’t recall other experts here saying that PWM is insufficient for power management. Power is, after all, the time derivative of work. And in fact most quality tubes have separate momentary and continuous power ratings. For example, my Reci 100W tube is rated to 130W momentary, 100W continuous.
I don’t want to get in a long debate here. But if you want to insist that I’m wrong, I’d ask for a citation for instantaneous power being a primary life-limiting factor, because I have read, as far as I know, everything posted to Maker Forums since I joined and definitely do not have the same impression you do about “including on this site”.