I've played a bit of rocket rush recently, played with 100% map rules but without blueprint import. I like that you have to build a large-ish base but runs only take a relatively short time. My pb is 1:15 so far and that's with very light planning and practice because I'm still evaluating strategies.

I think resources at default settings are not a good idea for rocket rush. If you only use the starter patches then you finish when you get all the copper you need out of the starter patch, which would drastically limit the time at which you can possibly finish and limit the depth of the category. But outposting also takes an uncomfortably large amount of time compared to how short the run is so it's not clear wether you'd save any time from it. It'd mostly come down to rerolling until you get a lucky copper patch and I'm opposed to that.

This maybe also applies for wave defense once you get fast enough.

EDIT: I have a replay but I changed my builds after the 1:15 run and haven't done a good run with the changes yet. I can give an outline of what I'm doing.

First goal is the mall, I'm going with two iron belts and half a belt of copper for this at the moment. Mall produces belt, inserters, miners (later changed to assemblers), pipes. Then build one steel smelter. Next goal is the first half of all mixed builds. I'm going with mixed builds that have integrated buffers for circuits and copper so the buffers can fill up a little while I set up oil and the second part of the mixed builds, which is what I do next. The mixed build is otherwise the nefrums mixed build, it outputs lds, blue circuits and modules, just slightly adapted because it has the buffers and I have no steel furnaces. I might change it to output rcus. Then I set up rocket fuel, electric engines, prod2s, rcus, concrete and finally the silo.

On how you determine which items you automate in which order: In general you should automate the one that costs the most to automate first. Which imo should be circuits, rocket fuel (including solid fuel), rcu, lds. If you automate something cheap first then this delays something costly so you have to build it larger to reach the same time and you pay more in total both in build time and mall. Slight tangent but if you compare nefrums any% design to older designs, he does this a lot: don't automate solid fuel until the end of the run, build steel for purple science after the circuit setup, order the rocket components as lds, rocket fuel, rcu in any% (because you need more lds).

This holds in general but there are exceptions of course. I'm still wondering wether it's better to get lds before rocket fuel and rcus in RR because they are more difficult to automate outside of a mixed build and the ratios work out a lot better - I'd have to figure out a way to use the copper otherwise.

There is also the question of which items you take with you. I did dive into how the prices are computed at some point. They first compute what I'll call the value of each item and then determine the price as value / log(value), so more valueable items have a slightly reduced price. The value is computed computed recursively from the ingredients with preset values assigned to raw ores and fluids (plus the logarithm of the crafting time but this almost doesnt matter because it's small). The thing you want to keep in mind is that valuable items are slightly cheaper, in particular items usually cost less than or equal to their ingredients.

I try to take just around enough of the basic items to last me until they are automated and automate basics asap. For example the first things I automate are stone for furnaces, power, an iron belt, transport belts. This cuts down the number of furnaces and belts that I need so I can take more higher tech items with me. I didn't do the calculation but I think that burner inserters are quite useful in this mode as they can be automated earlier and thus we need to buy less electric inserters. Generally I tend to get as much of the basics as I need and fill the rest with mining drills. Maybe when I get faster I'll take a few more robots or some batteries and advanced circuits to craft a second personal roboport but I'm not at the point where I need this yet.