Yes, to all, with the caveat that (4) will take a very long time as transient simulation proceeds incrementally, as discussed in another recent question. And consequently might not be very accurate (small numerical errors compounding over hundreds of cycles, etc.).
Note that (1) isn't very meaningful by itself, but you combine it with (2) to model the Thevenin or Norton equivalent source. In general, the antenna has a complex network equivalent, impedance and gain both varying with frequency, of course you can simply assume a given amplitude at any given frequency; the importance of impedance modeling, however, depends on what it's connected to.
(You can put the source almost anywhere in the antenna equivalent network, as long as it isn't changing the impedance thereof -- it can be a voltage source in series with any one component, or a current source between any two nodes.)
At some point, you should probably develop an understanding of what an RF port is, what incident and reflected waves are, impedance matching, tuning and all that. This will not come quickly, it's a deep topic, take it patiently -- but the value is immediate and huge: being able to break down a general N-component network into halves, joined by a port, through which current and voltage flow.
For example, if your receiver has a tuned input, and the tuning range is fairly wide, then it's probably safe to connect it to a narrower antenna, but do check that image bands haven't been created (resonances between antenna and receiver reactances, and the transmission line joining them, that don't exist otherwise), and which aren't rejected elsewhere in the system (because the antenna doesn't pick up those frequencies, or the receiver isn't sensitive to them). Doing this, requires modeling these sections themselves: SPICE provides a transmission line element which can be used, but the antenna you will likely have to measure yourself, and thus theory must be supported by practice: for example, use a signal generator and directional or reflectance bridge to measure the antenna's impedance.
Then apply the same definitions and principles, to build up a test jig in the simulator, and use that to match the impedance and gain of real networks you build.
Example:
I built this circuit in ye olde vacuum tubes for fun:
(leaving as links as this is an optional section, and to maintain copyright)
https://www.seventransistorlabs.com/Images/FMRadio3.jpg (note middle-bottom coils by the BNC connector)
https://www.seventransistorlabs.com/Images/FMRadio4.jpg (left coil)
using pluggable coils (0.1" header stock) in case I want to re-tune the set for different bands. The correpsonding model is:
https://www.seventransistorlabs.com/Images/Radio_IF_Filter.png
after some adaptation. The mixer (a 6J6 dual triode) has output impedance (plate-to-plate) of R9, I'm excluding modeling of the common mode (hence only one voltage source -- as this is a single-balanced mixer, the CMRR is relevant, but CM energy is also out-of-band so I kind of don't care), C10/C11 I believe are the plate capacitance and C8/C9 the trimmers, and L4/C12 is the output coupling network, a series resonant tank tuned to match to 50Ω with a modest Q. You can see from the plot, the response is supposed to be double-tuned (flat topped), which, worked out alright in practice I think; the measurement is:
https://www.seventransistorlabs.com/Images/FMRadio5.jpg
with a narrow marker tone at the center frequency which kind of hides the double-peaked response (or maybe it was a repeated-pole (single peak) at this stage, I don't recall).
It took several tries going back and forth, matching up predicted (design) and measured response, to arrive at a compromise where all component values look reasonable enough (within margin of mechanical measurements; use a calculator such as RF Inductance Calculator for Single‑Layer Helical Round‑Wire Coils
Serge Y. Stroobandt, ON4AA | hamwaves.com for accurate results on single-layer air-core helical windings) and give the predicted result. I used adjustments in the simulation to facilitate tuning the real thing -- particularly for steps that require more effort, like rewinding coils to make relatively large adjustments in inductance value and coupling factor.
One of the surprises I discovered through this project was, the grid (input) capacitance and resistance are much more severe than datasheets advertise -- in particular, the input capacitance of the 6688 RF preamp I used here, was more than double the datasheet value (which it turns out, is the cold, nonconducting figure only), and resistance was about half expected (they do actually give "hot" and conducting figures for these). In this case, there is a fascinating physical relevance to these figures: the capacitance increases because the grid is physically moving the electron cloud around the cathode; normally this is a conservative force (thus necessarily manifesting as inductance or capacitance in circuit, and the phase happens to be capacitive here), but any electrons lost from the cloud (that's current flow!) means force not returned to the grid -- resistance! So we are applying power to the grid, and, in a very true sense, performing work on the electron cloud and beam, modulating plate current and thus achieving amplification -- but at some limited power gain, not the infinite power gain you might assume from a nominally capacitive terminal (the grid).
The same isn't quite true of transistors, or more to the point, they're much "squishier" things in general (base/gate self-losses dominating over feedback phase shift, capacitance suffices to model feedback; etc.), but this does of course vary with type (RF transistors perform much better, and are more sensitive to their feedback characteristics, than GP/switching types). I just think it's interesting, given that tubes are such simple devices (visible construction, even), yet you can measure such a subtle mechanism (the collective motion of electrons) with such a strong effect in-circuit.
