Certain accounting practices specific to the software industry can also contribute to higher P/E ratios. For example, research and development (R&D) expenses for software companies are often treated as investments rather than expenses, which can artificially inflate earnings per share and further boost the P/E ratio.

I think R&D is treated as an expense which lowers earnings. This is a reason why software companies can support a higher PE.
If the R&D was treated as an investment, then the software company earnings would be higher, and for a given stock price the PE would be lower. Example, price $10, earnings before R&D $5, after subtraction of R&D as expense, earnings $4. PE 2.5.
If R&D treated as investment, Price $10, earnings before R&D $5, no subtraction of R&D. Earnings = $5 PE 2.

One of the reasons that it is difficult to compare historical PE ratios for the SP500 as the weighting of different industries within the SP500 has changed over time.

einvestingforbeginners.com - Rd spending as a percentage of revenue by industry

XOM has about $1B R&D and $25B capital expenditure, so earnings are reduced by $1B R&D and reduced for some depreciation
GOOG has about $39B R&D and $31B capital expenditure, so earnings are reduced by $39B R&D and reduced for some depreciation

Another factor for difficulty in comparing historical SP500 PEs is that stock holdings of companies such as BRK are now valued at mark to market pricing, so earnings are much more variable which means the PE ratio is much more variable

Craig