Jon,
After over 40 years on & off in the business, this is my current professional opinion. Keep in mind that other professionals from the industry, may disagree with what I value.
There are at least a couple of guys on the LSR site who also have a LOT of dyno experience. Perhaps they might offer their opinions as well. . . . . .
Most of the dyno work I have done has utilized SuperFlow (mostly 901 & some 902) equiptment. It is nice because everything is integrated into a system. What is important with any dyno is: day to day repeatability and accuracy. Without these, your development becomes meaningless. Accuracy to a standard allows you to compare any information you generate to other "properly calibrated" absorbers. There are people within the industry whose opinion is: that each absorber is a unique and individual unit that cannot be "accurately" compared to other units. When data collection was primitive, and somewhat subjective, this tended to be true. Now that data collection is computerized & more uniform, this tends to be less so. But, after having had 3 indentical cells & absorbers to utilize, I am also familiar with having 2 units agree and the other slightly higher or lower. This is why I value repeatability. I want to be able to reproduce the data that was generated on a particular pull +/- .3% I want to be able to "close the loop" in testing. If you do not know what I mean by "close the loop", buy and read "Dyno Testing & Tuning" by Harold Bettes & Bill Hancock.
Early in my career I used whatever dyno facility was available for the application I was working on. I used some Heenan & Froude absorbers and some Go-Power absorbers. All were manual data accumulation. Eventually, some were converted, more or less successfully. You need to make some hard choices about what features are important to the information/engineering you are going to gather. It may be more cost effective to replace your Go-Power with a good used SuperFlow, rather than trying to upgrade 1 system at a time or all 5 systems at once. Only you can make that kind of decision based on the cost of upgrading. I am currently working with an individual who has been upgrading his Go-Power, system by system over the past 2 years. His choices have all been economically driven, and still are. It has been a huge nightmare for him, and he still doesn't have the dyno he wants and needs. He builds engines professionally, and his product and output are suffering because of the constant changes to the dyno setup. . . . . .
When you are upgrading from whatever. . . You need to consider 5 issues:
1/ Load control
Going from manual to computerized/electronic. This may/probably will require modifications to the absorbers water inlet/outlet and/or the water supply/pressure pump.
This can get expensive.
2/ Torque measurement
Going from very old "dead weight scales" or old hydraulic transducers to electronic strain gages. Strain gages would be the most repeatable, but need to be properly
mounted and connected for "accuracy & repeatability". Some older absorbers may not adapt easily to current gages.
3/ Data accumulation
Going from "hand collected" to "computer collected". This is a matter of sensors & software. Shielding electronic components/sensors from ignition interference may
be an issue. In my opinion, more "data" (more collection channels) is always better. There is nothing more frustrating to analysis than to be missing vital input. . . . .
4/ Data processing/manipulation
Going from what is "collected" into "useful information." This is a matter of how the software calculates & produces information after the pull or run. Again, my opinion
is that, "more" is always better. Of course, if you do not use it, more info has little "value". A software system which allows you to upload to a desktop or laptop for
further analysis is a useful feature, in my opinion. Converting printouts, by hand, into spreadsheets/graphs/etc, is a giant pain in the A$$. And, if this is a business,
your time is money. . . . .
5/ Environment, ie, the dyno room/cell
Do not discount the value of controling the environment in which the engine(s) will be tested. In my opinion, this is absolutely necessary to have repeatability. This can
get expensive to provide ducts, blowers, tanks, pumps, etc, etc, etc. . . . .
If you are going to be serious about development, more info and better quality is always the way to go. If you just want to qualify or quantify numbers, a simpler system may serve your needs.
Fordboy