Unlevered and levered beta: multiplying the tax rate

Hi there, and thank you in advance:

In L1 the formulas for unlevering and levering the betas included the tax shield (1-t). Now, for L2, the formulas don’t include (1-t).

Wiley states the following: “The unlevering and relevering formulas presented here are the ones found in most textbooks (without the tax shield I assume), where it is assumed that debt levels are constant. The formulas presented in the curriculum do not multiply the D/E ratio by (1-t) as they assume that debt levels change over time. On the eam, look out for the assumptions that applies; otherwise use the formula from the curriculum”.

First of all, Wiley contradicts itself in this sentence (or am I missing something?).

Second, so, I do not understand, do we multiply by (1-t) when we assume debt levels change over time or when they do not change over time.

Other question, the pure play method (the one above) is always used and exclusively used for thinly traded companies, private businesses and projects? Or can it be used in for very liquid companies; and if so, why would we use that method? I suppose that the beta in the OLS regression has already in itself the D/E ratio of the company within, right?

My question is that if, for example, I want to compute the beta from Apple Inc. and do it by the ordinary least squares (OLS) regression and by the pure play method, so my question is the following:

What happens to the beta of Apple if its D/E ratio changes? In the OLS mehtod, would the company’s beta adjusts itself without the analyst having to make nothing to adjust it, so the market adjusts it automatically? If I was computing the beta by the pure play method I would need to adjust it right?

Thank you very much!!!