I agree with you if you look at Risk/Reward in isolation of other measures.
R:R is meaningless without also considering win rate, for example. If one’s strategy is based on a R:R of 1:1 then you know you also need a win rate in excess of 50% to generate an overall net gain.
Also, as you say, one cannot predict where price is going to reach. But neither should a stoploss level be based just on arithmetic whilst ignoring suitable levels where a stoploss order would be more sensible.
We design our strategies according to the price action characteristics of the instrument we select, the timeframe we use, the indicators and/or PA we prefer, and the size of our wallet, etc.
The strategy itself then defines the typical size of moves we are looking for and the distance to stoplosses that, if reached, would nullify the basis for having entered the trade.
When we define these two extremes we can then decide if the trade makes sense in R:R proportions.
On shorter timeframes, like 4H, 1H, and less, the R:R ratios defined by the strategy employed will probably be typically similar across all trades. For example, I doubt many people using 15min charts hold positions for several days.
So I would suggest that it is more the chosen strategy, and its overall win rate that defines the resultant typical R:R rather than using a pre-selected R:R to define the trades?
For example, I trade mainly off 1-hour charts with a win rate generally in the 65-75%range and the trades generated by my strategy typically have TPs and SLs defined by nearby S/R zones that inevitably form an R:R of around 1:1.
Afterall, if one looks for, say, an R:R of 1:4 on a 1-hour chart, then it will rarely succeed and the win rate will be very low. If, on the other hand, the TPs are sensible but the SL’s are too far away, say, 4:1, then the occasional hit on the SL will wipe out a lot of profit even though the win rate is fairly high.
Just some thoughts