FWIW, SpaceX has launched stages after full-length (3 minute plus) burns on the test stand, on a schedule that wouldn't allow time for that kind of teardown. They'll need to do something to clean the engines out eventually, but if several minutes of operation were enough to cause serious gunking, then between those test runs and the multiple runs of the grasshopper test vehicle for landing sequences, they've seen it and know how to deal with it.

(Incidentally, SpaceX is known for keeping some aspect of their operations trade secret, rather than patenting. So, if they had some solvent that dissolved the gunk and simplified the cleanup process, or some such similar trick, it's likely that no one outside the company would know about it.)

The first stage goes through at least one full mission duration burn and another ignition before finally launching. So they don't need to be reassembled completely between each firing. I assume they have been repeatedly firing engines for a while now to see how many relights they can get.

The hydrogen in the Space Shuttle Main Engine creates embrittlement in the alloys of the engine. [1][2] The launch sequence also involved running the engines at 104.5% of rated capacity. [3] Before Block II of the engines, the turbopumps needed to be disassembled and rebuilt between flights. [4]

I have no real idea how significant coking of the Merlin engine is, but the SSMEs were hardly "clean".

[1] https://en.wikipedia.org/wiki/Hydrogen_embrittlement [2] http://www.tms.org/Superalloys/10.7449/1991/Superalloys_1991... [3] http://www.interspacenews.com/FeatureArticle/tabid/130/Defau... [4] http://www.nasa.gov/sites/default/files/174534main_ssme.pdf

IIRC the 104% was additional performance they enabled through upgrades and design improvements beyond the original design not over-driving them.

That right. The 100% level was set at some point during the design process. Later they found ways to slightly increase the thrust. Rather than recalibrate all the numbers, they just kept using the original "100%" and called the new level "104%."

Basically, it's "104% of what the engineers circa 1975 thought the level would be," not 104% of what the hardware was rated for.

They did re-light the engines, curious how many times in a row they can pull that trick before contamination becomes a problem.