[Issue]: Tricky wildcard names resolution in tuples deconstruction

[silver-sd]

Current Implementation:

(var1, _) = (1, 2); //treated as the wildcard if the variable '_' is not declared

 int _;
 //...
 (var1, _) = (1, 2); //changes '_' variable. This is a potential bug(unintentional change).

Proposal:
Always treat this as the wildcard even if the variable '_' is declared:

(var1, _) = (1, 2)

If it is impossible(breaking change), at least generate a warning if the variable '_' is already declared. To reduce the probability of unintentional changes.

Like the current implementation of wildcards in 'out' parameters:

byte _ = 0;
F(out int _, out string _); //are always wildcards even if '_' is declared

The wildcard in 'switch' does not create any problems for using '_' as a variable too.

[svick]

It would be a breaking change, but it seems it's already being considered: #4460.