Editing flight models really involves three parts - the first part is the basic dimensions/weights/gear positions. Next is are the performance data. Lastly, is the supplemental data. Editing FMs is rather easy to do as long as you don't touch the Performance data. Editing FMs to be accurate is more difficult.
Doghouse utility is created by Julian “Codec” Onions who present time works for LeadPursuit (LP). FM Doghouse is readily available. If anyone wants a copy just PM Ranger822. However Ed_1 noted in this PMC Tactical forums post that Doghouse is outdated since SuperPAK 3 (SP3) code changes.
Aircraft FlightModels can be found inside Zip/simdata.zip/sim/ACDATA file.
If you unzip the simdata file to main directory (creating the inner sim folder there), all files there will override the same files existing in simdata.zip… means you can edit this files straight from the folder with no need to zip/unzip each time you change something and want to test it.
The .dat files themselves holds not only flight data but also lots of other data with variables. At least that's true for BMS based versions like RV and OF.
A good resource for FlightModels variables for F4 are located in the RP5 manual. You can also find some interesting data on the other variables introduced by BMS back then, here on this Frugals topic.
Ranger's Notes Regarding performance curves/data - this is where FM Doghouse is useful. It provides a graphic output of the curves - so you can compare relative performances - - This will allow anyone, to simply tweak the curves to more closely simulate a given flight performance desired. The main problem I have with the FM's in Falcon is with regard to having some sort of measurable standard or normalizing of performance values across all models. This may be a source of frustration for the user because sometimes the FMs attributed to a given aircraft may not be realistic and either the a/c has a better or worse performance than it actually should have. Further, without a measurable standard, relative performance to other aircraft, even if unrealistic or inaccurate, is not well determined. This can serve to explain why AI aircraft are either “really good” or “really bad” irrespective of the types of aircraft that are engaged in a dogfight/engagement. The problem gets worse over time as more and more people tamper with the FM's in an attempt to add more aircraft to the database, or “fix” existing problems.
Regarding FM variables. If you read thru the available literature from the forums or using RP5 manual you can get a rough idea of some of the variables. However, for those variable which are unclear, a way figure out the variables and what they do is to compare various models. You can tell right away by comparing FMs with their LOD counterparts there are significant differences. In most cases a  value means something is “off” and and “1” is either “on” or set to the first position. Most FMs variable only have a one or two positions. Experimentation is pretty easy. It would be nice to have some of the old BMS guys around to talk about FMs but they have long since retired so we are left to interpret what they have already written or make our own tests.
Testing FM changes. Cobra/RV allow you to unpack the Simdata folder in the zips folder and then by default it will use the unzipped folder first, per instructions above. This saves some time in switching out FMs that you want to test. A quick way to work testing is to simply rename your original file and add the new test file in its original location. In most cases the actype.lst file is large enough to accommodate most tests. However, if you want additional spaces you can easily add more airframe index numbers or simply identify the ones that are currently unused. If you want to add additional FMs you need to update the actype.lst file. It is a very simple file that lists the number of FMs subtract one from the actual number because it starts at “ zero”. Then each line starts the series 0,1,2,3,4 etc. You list the name of the file and the position it holds in the list equates to the number. Contact Ranger822 or Fafa for the actype.lst file for ITO. There is a little difference for Korea and all other theaters (less FM's). I don't know of any limit on the number of flight models. Had there been some centralized planning it might have been possible to allocate a couple of hundred FMs to each theater and then we could have shared a single simdata.zip. Just and idea.
In the near future I will post up a detailed FM walk some of the the variables I am familiar with and you can then cut and paste this “explanatory” FM for future reference.