is thorough and excellent!
In short, the 3He flux from the Sun is known and how deep those particles will embed themselves in the regolith surface and the rate of regolith heating, cooling, and overturn by meteorites can all be modeled.
So the methods outlined do not actually detect 3He at all, they just assume that it must be there.
Here I'll address the "could" part of the question
How could the presence of 3He be detected on the lunar surface?
but I have not found published explanations on how presence and concentration could or has be detected.
If I wanted to find direct evidence of 3He on the Moon and try to make a quantitative measurement of it, I would look for ~21 MeV gamma rays produced by thermal neutron capture of 3He.
From Energy Levels of Light Nuclei; A=4 (1992) starting with reaction #14: 3He(n, γ)4He, Qm = 20.578 and
Table 4.14: Measured values for the thermal neutron capture cross
section for the 3H(n, γ) 4He reaction
we can see that this reaction leading to a distinct high-energy gamma ray (or it's detailed-balance inverse) has astrophysical implications and for diagnostics of fusion reactors so has been studied extensively.
The thermal neutron radiative capture cross-section is about 60 micro-barns (yes) which is not big but it's not vanishingly small either.
Water has been successfully(?) prospected on the moon via energetic neutron scattering by protons (hydrogen) but I don't know the thermal neutron flux at the Moon's surface well enough to estimate the gamma ray rate expected in orbit.
Since the coverage is expected to be widespread, it won't be any stronger if you are 1 meter above the Moon than if you were in orbit 100 km above it, for the same reason that a wall doesn't get brighter when you walk towards it. (cf. etendue)
Now, a detector that can capture a 20.6 MeV gamma ray and determine its energy precisely will be a bit of a challenge. A big old NaI scintillator would work, but it would have to be pretty big to contain the full shower produced by such a high energy gamma ray. For this energy one might look into the lower resolution BGO (bismuth germinate) scintillator detectors, smaller because bismuth and germanium have a lot more electrons and higher nuclear charge for shower-stopping, but perhaps just as heavy.
If the gamma ray background is too high, then you'll need a gamma ray detector with much higher resolution, so that a weak peak might stand out more sharply. A germanium detector (a big single crystal germanium reverse-biased diode) would provide much higher resolution, but again it would have to be exceedingly large to contain all the energy of a 20.6 MeV gamma ray much of the time.
Source: Energy Levels of Light Nuclei; A=4 (1992)
Fig. 2: The energy levels of 4He are plotted on a vertical scale giving the c.m. energy, in MeV, relative to its ground state. Horizontal lines representing the levels are labeled by the level energies and values of total angular momentum, parity, and isospin (Jπ, T ). Also shown are threshold energies and typical thin-target excitation functions for some of the reactions in the 4He system. See Fig. 1 for further details about notation and Table 4.3 for more information about the levels, including partial and total widths.